* [PATCH 0/7] Latest NetLabel patch for 2.6.19
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt
I am posting this patchset for consideration and inclusion into the 2.6.19
kernel, it is against 2.6.18-rc1.
This patchset introduces NetLabel, a implementation of explicit packet
labeling (i.e. CIPSO), to the Linux kernel. NetLabel has been designed to
have as minimal an impact on the base networking stack as possible; this
includes both code changes as well as performance. I, as well as many others
who have posted to various lists on earlier NetLabel patches, believe that an
interoperable form of labeled networking is important for Linux's success in
the Trusted OS arena currently being dominated by commercial UNIX systems.
DaveM, I know you have previously posted that you feel CIPSO does not belong
in the Linux kernel on principle, however, I'm hoping the arguments posted
in response have softened your position ...
Earlier versions of this patchset have been posted to the netdev, SELinux,
LSM and RH-LSPP mailing lists over the past couple of months. It now contains
several rounds of comments and has been tested on a variety of architectures
by people on the RH-LSPP mailing list over the course of the last several
weeks.
If accepted into the mainline kernel, both HP and myself pledge to maintain
this code.
- Notes on Performance
This past week there was a thread on the RH-LSPP list where the performance of
the NetLabel patch was measured and discussed using the 2.6.17 kernel. A copy
of the discussion can be found here:
* http://www.redhat.com/archives/redhat-lspp/2006-July/msg00063.html
With the conclusion being that performance should not be an issue.
Unfortunately the vanilla 2.6.18-rc1 kernel has problems on the two machines
I use for performance testing so I am not currently able to update the
NetLabel performance numbers for 2.6.18-rc1.
- Notes on Interoperability Testing
The NetLabel CIPSO implementation has been tested against Trusted Solaris and
HP-UX CMW without problems.
- Instructions for Testing
For those of you wishing to test this patchset you will need the latest
release of the netlabel_tools tarball found here:
* http://free.linux.hp.com/~pmoore/projects/linux_cipso
You also may want to make use of the "toy policy module" for SELinux which has
been posted to the RH-LSPP mailing list, the archived message can be found
here:
* http://www.redhat.com/archives/redhat-lspp/2006-June/msg00243.html
Thanks.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 0/7] Latest NetLabel patch for 2.6.19
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt
I am posting this patchset for consideration and inclusion into the 2.6.19
kernel, it is against 2.6.18-rc1.
This patchset introduces NetLabel, a implementation of explicit packet
labeling (i.e. CIPSO), to the Linux kernel. NetLabel has been designed to
have as minimal an impact on the base networking stack as possible; this
includes both code changes as well as performance. I, as well as many others
who have posted to various lists on earlier NetLabel patches, believe that an
interoperable form of labeled networking is important for Linux's success in
the Trusted OS arena currently being dominated by commercial UNIX systems.
DaveM, I know you have previously posted that you feel CIPSO does not belong
in the Linux kernel on principle, however, I'm hoping the arguments posted
in response have softened your position ...
Earlier versions of this patchset have been posted to the netdev, SELinux,
LSM and RH-LSPP mailing lists over the past couple of months. It now contains
several rounds of comments and has been tested on a variety of architectures
by people on the RH-LSPP mailing list over the course of the last several
weeks.
If accepted into the mainline kernel, both HP and myself pledge to maintain
this code.
- Notes on Performance
This past week there was a thread on the RH-LSPP list where the performance of
the NetLabel patch was measured and discussed using the 2.6.17 kernel. A copy
of the discussion can be found here:
* http://www.redhat.com/archives/redhat-lspp/2006-July/msg00063.html
With the conclusion being that performance should not be an issue.
Unfortunately the vanilla 2.6.18-rc1 kernel has problems on the two machines
I use for performance testing so I am not currently able to update the
NetLabel performance numbers for 2.6.18-rc1.
- Notes on Interoperability Testing
The NetLabel CIPSO implementation has been tested against Trusted Solaris and
HP-UX CMW without problems.
- Instructions for Testing
For those of you wishing to test this patchset you will need the latest
release of the netlabel_tools tarball found here:
* http://free.linux.hp.com/~pmoore/projects/linux_cipso
You also may want to make use of the "toy policy module" for SELinux which has
been posted to the RH-LSPP mailing list, the archived message can be found
here:
* http://www.redhat.com/archives/redhat-lspp/2006-June/msg00243.html
Thanks.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 1/7] NetLabel: documentation
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-documentation --]
[-- Type: text/plain, Size: 41871 bytes --]
Documentation for the NetLabel system, this includes a basic overview of how
NetLabel works, how LSM developers can integrate it into their favorite
LSM, as well as documentation on the CIPSO related sysctl variables. Also,
due to the difficulty of finding expired IETF drafts, I am including the IETF
CIPSO draft that is the basis of the NetLabel CIPSO implementation.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
CREDITS | 7
Documentation/00-INDEX | 2
Documentation/netlabel/00-INDEX | 10
Documentation/netlabel/cipso_ipv4.txt | 48
Documentation/netlabel/draft-ietf-cipso-ipsecurity-01.txt | 791 ++++++++++++++
Documentation/netlabel/introduction.txt | 53
Documentation/netlabel/lsm_interface.txt | 47
Documentation/networking/ip-sysctl.txt | 28
8 files changed, 986 insertions(+)
Index: linux-2.6.18-rc1/CREDITS
===================================================================
--- linux-2.6.18-rc1.orig/CREDITS
+++ linux-2.6.18-rc1/CREDITS
@@ -2384,6 +2384,13 @@ N: Thomas Molina
E: tmolina@cablespeed.com
D: bug fixes, documentation, minor hackery
+N: Paul Moore
+E: paul.moore@hp.com
+D: NetLabel author
+S: Hewlett-Packard
+S: 110 Spit Brook Road
+S: Nashua, NH 03062
+
N: James Morris
E: jmorris@namei.org
W: http://namei.org/
Index: linux-2.6.18-rc1/Documentation/00-INDEX
===================================================================
--- linux-2.6.18-rc1.orig/Documentation/00-INDEX
+++ linux-2.6.18-rc1/Documentation/00-INDEX
@@ -184,6 +184,8 @@ mtrr.txt
- how to use PPro Memory Type Range Registers to increase performance.
nbd.txt
- info on a TCP implementation of a network block device.
+netlabel/
+ - directory with information on the NetLabel subsystem.
networking/
- directory with info on various aspects of networking with Linux.
nfsroot.txt
Index: linux-2.6.18-rc1/Documentation/netlabel/00-INDEX
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/00-INDEX
@@ -0,0 +1,10 @@
+00-INDEX
+ - this file.
+cipso_ipv4.txt
+ - documentation on the IPv4 CIPSO protocol engine.
+draft-ietf-cipso-ipsecurity-01.txt
+ - IETF draft of the CIPSO protocol, dated 16 July 1992.
+introduction.txt
+ - NetLabel introduction, READ THIS FIRST.
+lsm_interface.txt
+ - documentation on the NetLabel kernel security module API.
Index: linux-2.6.18-rc1/Documentation/netlabel/cipso_ipv4.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/cipso_ipv4.txt
@@ -0,0 +1,48 @@
+NetLabel CIPSO/IPv4 Protocol Engine
+==============================================================================
+Paul Moore, paul.moore@hp.com
+
+May 17, 2006
+
+ * Overview
+
+The NetLabel CIPSO/IPv4 protocol engine is based on the IETF Commercial IP
+Security Option (CIPSO) draft from July 16, 1992. A copy of this draft can be
+found in this directory, consult '00-INDEX' for the filename. While the IETF
+draft never made it to an RFC standard it has become a de-facto standard for
+labeled networking and is used in many trusted operating systems.
+
+ * Outbound Packet Processing
+
+The CIPSO/IPv4 protocol engine applies the CIPSO IP option to packets by
+adding the CIPSO label to the socket. This causes all packets leaving the
+system through the socket to have the CIPSO IP option applied. The socket's
+CIPSO label can be changed at any point in time, however, it is recommended
+that it is set upon the socket's creation. The LSM can set the socket's CIPSO
+label by using the NetLabel security module API; if the NetLabel "domain" is
+configured to use CIPSO for packet labeling then a CIPSO IP option will be
+generated and attached to the socket.
+
+ * Inbound Packet Processing
+
+The CIPSO/IPv4 protocol engine validates every CIPSO IP option it finds at the
+IP layer without any special handling required by the LSM. However, in order
+to decode and translate the CIPSO label on the packet the LSM must use the
+NetLabel security module API to extract the security attributes of the packet.
+This is typically done at the socket layer using the 'socket_sock_rcv_skb()'
+LSM hook.
+
+ * Label Translation
+
+The CIPSO/IPv4 protocol engine contains a mechanism to translate CIPSO security
+attributes such as sensitivity level and category to values which are
+appropriate for the host. These mappings are defined as part of a CIPSO
+Domain Of Interpretation (DOI) definition and are configured through the
+NetLabel user space communication layer. Each DOI definition can have a
+different security attribute mapping table.
+
+ * Label Translation Cache
+
+The NetLabel system provides a framework for caching security attribute
+mappings from the network labels to the corresponding LSM identifiers. The
+CIPSO/IPv4 protocol engine supports this caching mechanism.
Index: linux-2.6.18-rc1/Documentation/netlabel/draft-ietf-cipso-ipsecurity-01.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/draft-ietf-cipso-ipsecurity-01.txt
@@ -0,0 +1,791 @@
+IETF CIPSO Working Group
+16 July, 1992
+
+
+
+ COMMERCIAL IP SECURITY OPTION (CIPSO 2.2)
+
+
+
+1. Status
+
+This Internet Draft provides the high level specification for a Commercial
+IP Security Option (CIPSO). This draft reflects the version as approved by
+the CIPSO IETF Working Group. Distribution of this memo is unlimited.
+
+This document is an Internet Draft. Internet Drafts are working documents
+of the Internet Engineering Task Force (IETF), its Areas, and its Working
+Groups. Note that other groups may also distribute working documents as
+Internet Drafts.
+
+Internet Drafts are draft documents valid for a maximum of six months.
+Internet Drafts may be updated, replaced, or obsoleted by other documents
+at any time. It is not appropriate to use Internet Drafts as reference
+material or to cite them other than as a "working draft" or "work in
+progress."
+
+Please check the I-D abstract listing contained in each Internet Draft
+directory to learn the current status of this or any other Internet Draft.
+
+
+
+
+2. Background
+
+Currently the Internet Protocol includes two security options. One of
+these options is the DoD Basic Security Option (BSO) (Type 130) which allows
+IP datagrams to be labeled with security classifications. This option
+provides sixteen security classifications and a variable number of handling
+restrictions. To handle additional security information, such as security
+categories or compartments, another security option (Type 133) exists and
+is referred to as the DoD Extended Security Option (ESO). The values for
+the fixed fields within these two options are administered by the Defense
+Information Systems Agency (DISA).
+
+Computer vendors are now building commercial operating systems with
+mandatory access controls and multi-level security. These systems are
+no longer built specifically for a particular group in the defense or
+intelligence communities. They are generally available commercial systems
+for use in a variety of government and civil sector environments.
+
+The small number of ESO format codes can not support all the possible
+applications of a commercial security option. The BSO and ESO were
+designed to only support the United States DoD. CIPSO has been designed
+to support multiple security policies. This Internet Draft provides the
+format and procedures required to support a Mandatory Access Control
+security policy. Support for additional security policies shall be
+defined in future RFCs.
+
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 1]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+
+3. CIPSO Format
+
+Option type: 134 (Class 0, Number 6, Copy on Fragmentation)
+Option length: Variable
+
+This option permits security related information to be passed between
+systems within a single Domain of Interpretation (DOI). A DOI is a
+collection of systems which agree on the meaning of particular values
+in the security option. An authority that has been assigned a DOI
+identifier will define a mapping between appropriate CIPSO field values
+and their human readable equivalent. This authority will distribute that
+mapping to hosts within the authority's domain. These mappings may be
+sensitive, therefore a DOI authority is not required to make these
+mappings available to anyone other than the systems that are included in
+the DOI.
+
+This option MUST be copied on fragmentation. This option appears at most
+once in a datagram. All multi-octet fields in the option are defined to be
+transmitted in network byte order. The format of this option is as follows:
+
++----------+----------+------//------+-----------//---------+
+| 10000110 | LLLLLLLL | DDDDDDDDDDDD | TTTTTTTTTTTTTTTTTTTT |
++----------+----------+------//------+-----------//---------+
+
+ TYPE=134 OPTION DOMAIN OF TAGS
+ LENGTH INTERPRETATION
+
+
+ Figure 1. CIPSO Format
+
+
+3.1 Type
+
+This field is 1 octet in length. Its value is 134.
+
+
+3.2 Length
+
+This field is 1 octet in length. It is the total length of the option
+including the type and length fields. With the current IP header length
+restriction of 40 octets the value of this field MUST not exceed 40.
+
+
+3.3 Domain of Interpretation Identifier
+
+This field is an unsigned 32 bit integer. The value 0 is reserved and MUST
+not appear as the DOI identifier in any CIPSO option. Implementations
+should assume that the DOI identifier field is not aligned on any particular
+byte boundary.
+
+To conserve space in the protocol, security levels and categories are
+represented by numbers rather than their ASCII equivalent. This requires
+a mapping table within CIPSO hosts to map these numbers to their
+corresponding ASCII representations. Non-related groups of systems may
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 2]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+have their own unique mappings. For example, one group of systems may
+use the number 5 to represent Unclassified while another group may use the
+number 1 to represent that same security level. The DOI identifier is used
+to identify which mapping was used for the values within the option.
+
+
+3.4 Tag Types
+
+A common format for passing security related information is necessary
+for interoperability. CIPSO uses sets of "tags" to contain the security
+information relevant to the data in the IP packet. Each tag begins with
+a tag type identifier followed by the length of the tag and ends with the
+actual security information to be passed. All multi-octet fields in a tag
+are defined to be transmitted in network byte order. Like the DOI
+identifier field in the CIPSO header, implementations should assume that
+all tags, as well as fields within a tag, are not aligned on any particular
+octet boundary. The tag types defined in this document contain alignment
+bytes to assist alignment of some information, however alignment can not
+be guaranteed if CIPSO is not the first IP option.
+
+CIPSO tag types 0 through 127 are reserved for defining standard tag
+formats. Their definitions will be published in RFCs. Tag types whose
+identifiers are greater than 127 are defined by the DOI authority and may
+only be meaningful in certain Domains of Interpretation. For these tag
+types, implementations will require the DOI identifier as well as the tag
+number to determine the security policy and the format associated with the
+tag. Use of tag types above 127 are restricted to closed networks where
+interoperability with other networks will not be an issue. Implementations
+that support a tag type greater than 127 MUST support at least one DOI that
+requires only tag types 1 to 127.
+
+Tag type 0 is reserved. Tag types 1, 2, and 5 are defined in this
+Internet Draft. Types 3 and 4 are reserved for work in progress.
+The standard format for all current and future CIPSO tags is shown below:
+
++----------+----------+--------//--------+
+| TTTTTTTT | LLLLLLLL | IIIIIIIIIIIIIIII |
++----------+----------+--------//--------+
+ TAG TAG TAG
+ TYPE LENGTH INFORMATION
+
+ Figure 2: Standard Tag Format
+
+In the three tag types described in this document, the length and count
+restrictions are based on the current IP limitation of 40 octets for all
+IP options. If the IP header is later expanded, then the length and count
+restrictions specified in this document may increase to use the full area
+provided for IP options.
+
+
+3.4.1 Tag Type Classes
+
+Tag classes consist of tag types that have common processing requirements
+and support the same security policy. The three tags defined in this
+Internet Draft belong to the Mandatory Access Control (MAC) Sensitivity
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 3]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+class and support the MAC Sensitivity security policy.
+
+
+3.4.2 Tag Type 1
+
+This is referred to as the "bit-mapped" tag type. Tag type 1 is included
+in the MAC Sensitivity tag type class. The format of this tag type is as
+follows:
+
++----------+----------+----------+----------+--------//---------+
+| 00000001 | LLLLLLLL | 00000000 | LLLLLLLL | CCCCCCCCCCCCCCCCC |
++----------+----------+----------+----------+--------//---------+
+
+ TAG TAG ALIGNMENT SENSITIVITY BIT MAP OF
+ TYPE LENGTH OCTET LEVEL CATEGORIES
+
+ Figure 3. Tag Type 1 Format
+
+
+3.4.2.1 Tag Type
+
+This field is 1 octet in length and has a value of 1.
+
+
+3.4.2.2 Tag Length
+
+This field is 1 octet in length. It is the total length of the tag type
+including the type and length fields. With the current IP header length
+restriction of 40 bytes the value within this field is between 4 and 34.
+
+
+3.4.2.3 Alignment Octet
+
+This field is 1 octet in length and always has the value of 0. Its purpose
+is to align the category bitmap field on an even octet boundary. This will
+speed many implementations including router implementations.
+
+
+3.4.2.4 Sensitivity Level
+
+This field is 1 octet in length. Its value is from 0 to 255. The values
+are ordered with 0 being the minimum value and 255 representing the maximum
+value.
+
+
+3.4.2.5 Bit Map of Categories
+
+The length of this field is variable and ranges from 0 to 30 octets. This
+provides representation of categories 0 to 239. The ordering of the bits
+is left to right or MSB to LSB. For example category 0 is represented by
+the most significant bit of the first byte and category 15 is represented
+by the least significant bit of the second byte. Figure 4 graphically
+shows this ordering. Bit N is binary 1 if category N is part of the label
+for the datagram, and bit N is binary 0 if category N is not part of the
+label. Except for the optimized tag 1 format described in the next section,
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 4]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+minimal encoding SHOULD be used resulting in no trailing zero octets in the
+category bitmap.
+
+ octet 0 octet 1 octet 2 octet 3 octet 4 octet 5
+ XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX . . .
+bit 01234567 89111111 11112222 22222233 33333333 44444444
+number 012345 67890123 45678901 23456789 01234567
+
+ Figure 4. Ordering of Bits in Tag 1 Bit Map
+
+
+3.4.2.6 Optimized Tag 1 Format
+
+Routers work most efficiently when processing fixed length fields. To
+support these routers there is an optimized form of tag type 1. The format
+does not change. The only change is to the category bitmap which is set to
+a constant length of 10 octets. Trailing octets required to fill out the 10
+octets are zero filled. Ten octets, allowing for 80 categories, was chosen
+because it makes the total length of the CIPSO option 20 octets. If CIPSO
+is the only option then the option will be full word aligned and additional
+filler octets will not be required.
+
+
+3.4.3 Tag Type 2
+
+This is referred to as the "enumerated" tag type. It is used to describe
+large but sparsely populated sets of categories. Tag type 2 is in the MAC
+Sensitivity tag type class. The format of this tag type is as follows:
+
++----------+----------+----------+----------+-------------//-------------+
+| 00000010 | LLLLLLLL | 00000000 | LLLLLLLL | CCCCCCCCCCCCCCCCCCCCCCCCCC |
++----------+----------+----------+----------+-------------//-------------+
+
+ TAG TAG ALIGNMENT SENSITIVITY ENUMERATED
+ TYPE LENGTH OCTET LEVEL CATEGORIES
+
+ Figure 5. Tag Type 2 Format
+
+
+3.4.3.1 Tag Type
+
+This field is one octet in length and has a value of 2.
+
+
+3.4.3.2 Tag Length
+
+This field is 1 octet in length. It is the total length of the tag type
+including the type and length fields. With the current IP header length
+restriction of 40 bytes the value within this field is between 4 and 34.
+
+
+3.4.3.3 Alignment Octet
+
+This field is 1 octet in length and always has the value of 0. Its purpose
+is to align the category field on an even octet boundary. This will
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 5]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+speed many implementations including router implementations.
+
+
+3.4.3.4 Sensitivity Level
+
+This field is 1 octet in length. Its value is from 0 to 255. The values
+are ordered with 0 being the minimum value and 255 representing the
+maximum value.
+
+
+3.4.3.5 Enumerated Categories
+
+In this tag, categories are represented by their actual value rather than
+by their position within a bit field. The length of each category is 2
+octets. Up to 15 categories may be represented by this tag. Valid values
+for categories are 0 to 65534. Category 65535 is not a valid category
+value. The categories MUST be listed in ascending order within the tag.
+
+
+3.4.4 Tag Type 5
+
+This is referred to as the "range" tag type. It is used to represent
+labels where all categories in a range, or set of ranges, are included
+in the sensitivity label. Tag type 5 is in the MAC Sensitivity tag type
+class. The format of this tag type is as follows:
+
++----------+----------+----------+----------+------------//-------------+
+| 00000101 | LLLLLLLL | 00000000 | LLLLLLLL | Top/Bottom | Top/Bottom |
++----------+----------+----------+----------+------------//-------------+
+
+ TAG TAG ALIGNMENT SENSITIVITY CATEGORY RANGES
+ TYPE LENGTH OCTET LEVEL
+
+ Figure 6. Tag Type 5 Format
+
+
+3.4.4.1 Tag Type
+
+This field is one octet in length and has a value of 5.
+
+
+3.4.4.2 Tag Length
+
+This field is 1 octet in length. It is the total length of the tag type
+including the type and length fields. With the current IP header length
+restriction of 40 bytes the value within this field is between 4 and 34.
+
+
+3.4.4.3 Alignment Octet
+
+This field is 1 octet in length and always has the value of 0. Its purpose
+is to align the category range field on an even octet boundary. This will
+speed many implementations including router implementations.
+
+
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 6]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+3.4.4.4 Sensitivity Level
+
+This field is 1 octet in length. Its value is from 0 to 255. The values
+are ordered with 0 being the minimum value and 255 representing the maximum
+value.
+
+
+3.4.4.5 Category Ranges
+
+A category range is a 4 octet field comprised of the 2 octet index of the
+highest numbered category followed by the 2 octet index of the lowest
+numbered category. These range endpoints are inclusive within the range of
+categories. All categories within a range are included in the sensitivity
+label. This tag may contain a maximum of 7 category pairs. The bottom
+category endpoint for the last pair in the tag MAY be omitted and SHOULD be
+assumed to be 0. The ranges MUST be non-overlapping and be listed in
+descending order. Valid values for categories are 0 to 65534. Category
+65535 is not a valid category value.
+
+
+3.4.5 Minimum Requirements
+
+A CIPSO implementation MUST be capable of generating at least tag type 1 in
+the non-optimized form. In addition, a CIPSO implementation MUST be able
+to receive any valid tag type 1 even those using the optimized tag type 1
+format.
+
+
+4. Configuration Parameters
+
+The configuration parameters defined below are required for all CIPSO hosts,
+gateways, and routers that support multiple sensitivity labels. A CIPSO
+host is defined to be the origination or destination system for an IP
+datagram. A CIPSO gateway provides IP routing services between two or more
+IP networks and may be required to perform label translations between
+networks. A CIPSO gateway may be an enhanced CIPSO host or it may just
+provide gateway services with no end system CIPSO capabilities. A CIPSO
+router is a dedicated IP router that routes IP datagrams between two or more
+IP networks.
+
+An implementation of CIPSO on a host MUST have the capability to reject a
+datagram for reasons that the information contained can not be adequately
+protected by the receiving host or if acceptance may result in violation of
+the host or network security policy. In addition, a CIPSO gateway or router
+MUST be able to reject datagrams going to networks that can not provide
+adequate protection or may violate the network's security policy. To
+provide this capability the following minimal set of configuration
+parameters are required for CIPSO implementations:
+
+HOST_LABEL_MAX - This parameter contains the maximum sensitivity label that
+a CIPSO host is authorized to handle. All datagrams that have a label
+greater than this maximum MUST be rejected by the CIPSO host. This
+parameter does not apply to CIPSO gateways or routers. This parameter need
+not be defined explicitly as it can be implicitly derived from the
+PORT_LABEL_MAX parameters for the associated interfaces.
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 7]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+
+HOST_LABEL_MIN - This parameter contains the minimum sensitivity label that
+a CIPSO host is authorized to handle. All datagrams that have a label less
+than this minimum MUST be rejected by the CIPSO host. This parameter does
+not apply to CIPSO gateways or routers. This parameter need not be defined
+explicitly as it can be implicitly derived from the PORT_LABEL_MIN
+parameters for the associated interfaces.
+
+PORT_LABEL_MAX - This parameter contains the maximum sensitivity label for
+all datagrams that may exit a particular network interface port. All
+outgoing datagrams that have a label greater than this maximum MUST be
+rejected by the CIPSO system. The label within this parameter MUST be
+less than or equal to the label within the HOST_LABEL_MAX parameter. This
+parameter does not apply to CIPSO hosts that support only one network port.
+
+PORT_LABEL_MIN - This parameter contains the minimum sensitivity label for
+all datagrams that may exit a particular network interface port. All
+outgoing datagrams that have a label less than this minimum MUST be
+rejected by the CIPSO system. The label within this parameter MUST be
+greater than or equal to the label within the HOST_LABEL_MIN parameter.
+This parameter does not apply to CIPSO hosts that support only one network
+port.
+
+PORT_DOI - This parameter is used to assign a DOI identifier value to a
+particular network interface port. All CIPSO labels within datagrams
+going out this port MUST use the specified DOI identifier. All CIPSO
+hosts and gateways MUST support either this parameter, the NET_DOI
+parameter, or the HOST_DOI parameter.
+
+NET_DOI - This parameter is used to assign a DOI identifier value to a
+particular IP network address. All CIPSO labels within datagrams destined
+for the particular IP network MUST use the specified DOI identifier. All
+CIPSO hosts and gateways MUST support either this parameter, the PORT_DOI
+parameter, or the HOST_DOI parameter.
+
+HOST_DOI - This parameter is used to assign a DOI identifier value to a
+particular IP host address. All CIPSO labels within datagrams destined for
+the particular IP host will use the specified DOI identifier. All CIPSO
+hosts and gateways MUST support either this parameter, the PORT_DOI
+parameter, or the NET_DOI parameter.
+
+This list represents the minimal set of configuration parameters required
+to be compliant. Implementors are encouraged to add to this list to
+provide enhanced functionality and control. For example, many security
+policies may require both incoming and outgoing datagrams be checked against
+the port and host label ranges.
+
+
+4.1 Port Range Parameters
+
+The labels represented by the PORT_LABEL_MAX and PORT_LABEL_MIN parameters
+MAY be in CIPSO or local format. Some CIPSO systems, such as routers, may
+want to have the range parameters expressed in CIPSO format so that incoming
+labels do not have to be converted to a local format before being compared
+against the range. If multiple DOIs are supported by one of these CIPSO
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 8]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+systems then multiple port range parameters would be needed, one set for
+each DOI supported on a particular port.
+
+The port range will usually represent the total set of labels that may
+exist on the logical network accessed through the corresponding network
+interface. It may, however, represent a subset of these labels that are
+allowed to enter the CIPSO system.
+
+
+4.2 Single Label CIPSO Hosts
+
+CIPSO implementations that support only one label are not required to
+support the parameters described above. These limited implementations are
+only required to support a NET_LABEL parameter. This parameter contains
+the CIPSO label that may be inserted in datagrams that exit the host. In
+addition, the host MUST reject any incoming datagram that has a label which
+is not equivalent to the NET_LABEL parameter.
+
+
+5. Handling Procedures
+
+This section describes the processing requirements for incoming and
+outgoing IP datagrams. Just providing the correct CIPSO label format
+is not enough. Assumptions will be made by one system on how a
+receiving system will handle the CIPSO label. Wrong assumptions may
+lead to non-interoperability or even a security incident. The
+requirements described below represent the minimal set needed for
+interoperability and that provide users some level of confidence.
+Many other requirements could be added to increase user confidence,
+however at the risk of restricting creativity and limiting vendor
+participation.
+
+
+5.1 Input Procedures
+
+All datagrams received through a network port MUST have a security label
+associated with them, either contained in the datagram or assigned to the
+receiving port. Without this label the host, gateway, or router will not
+have the information it needs to make security decisions. This security
+label will be obtained from the CIPSO if the option is present in the
+datagram. See section 4.1.2 for handling procedures for unlabeled
+datagrams. This label will be compared against the PORT (if appropriate)
+and HOST configuration parameters defined in section 3.
+
+If any field within the CIPSO option, such as the DOI identifier, is not
+recognized the IP datagram is discarded and an ICMP "parameter problem"
+(type 12) is generated and returned. The ICMP code field is set to "bad
+parameter" (code 0) and the pointer is set to the start of the CIPSO field
+that is unrecognized.
+
+If the contents of the CIPSO are valid but the security label is
+outside of the configured host or port label range, the datagram is
+discarded and an ICMP "destination unreachable" (type 3) is generated
+and returned. The code field of the ICMP is set to "communication with
+destination network administratively prohibited" (code 9) or to
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 9]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+"communication with destination host administratively prohibited"
+(code 10). The value of the code field used is dependent upon whether
+the originator of the ICMP message is acting as a CIPSO host or a CIPSO
+gateway. The recipient of the ICMP message MUST be able to handle either
+value. The same procedure is performed if a CIPSO can not be added to an
+IP packet because it is too large to fit in the IP options area.
+
+If the error is triggered by receipt of an ICMP message, the message
+is discarded and no response is permitted (consistent with general ICMP
+processing rules).
+
+
+5.1.1 Unrecognized tag types
+
+The default condition for any CIPSO implementation is that an
+unrecognized tag type MUST be treated as a "parameter problem" and
+handled as described in section 4.1. A CIPSO implementation MAY allow
+the system administrator to identify tag types that may safely be
+ignored. This capability is an allowable enhancement, not a
+requirement.
+
+
+5.1.2 Unlabeled Packets
+
+A network port may be configured to not require a CIPSO label for all
+incoming datagrams. For this configuration a CIPSO label must be
+assigned to that network port and associated with all unlabeled IP
+datagrams. This capability might be used for single level networks or
+networks that have CIPSO and non-CIPSO hosts and the non-CIPSO hosts
+all operate at the same label.
+
+If a CIPSO option is required and none is found, the datagram is
+discarded and an ICMP "parameter problem" (type 12) is generated and
+returned to the originator of the datagram. The code field of the ICMP
+is set to "option missing" (code 1) and the ICMP pointer is set to 134
+(the value of the option type for the missing CIPSO option).
+
+
+5.2 Output Procedures
+
+A CIPSO option MUST appear only once in a datagram. Only one tag type
+from the MAC Sensitivity class MAY be included in a CIPSO option. Given
+the current set of defined tag types, this means that CIPSO labels at
+first will contain only one tag.
+
+All datagrams leaving a CIPSO system MUST meet the following condition:
+
+ PORT_LABEL_MIN <= CIPSO label <= PORT_LABEL_MAX
+
+If this condition is not satisfied the datagram MUST be discarded.
+If the CIPSO system only supports one port, the HOST_LABEL_MIN and the
+HOST_LABEL_MAX parameters MAY be substituted for the PORT parameters in
+the above condition.
+
+The DOI identifier to be used for all outgoing datagrams is configured by
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 10]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+the administrator. If port level DOI identifier assignment is used, then
+the PORT_DOI configuration parameter MUST contain the DOI identifier to
+use. If network level DOI assignment is used, then the NET_DOI parameter
+MUST contain the DOI identifier to use. And if host level DOI assignment
+is employed, then the HOST_DOI parameter MUST contain the DOI identifier
+to use. A CIPSO implementation need only support one level of DOI
+assignment.
+
+
+5.3 DOI Processing Requirements
+
+A CIPSO implementation MUST support at least one DOI and SHOULD support
+multiple DOIs. System and network administrators are cautioned to
+ensure that at least one DOI is common within an IP network to allow for
+broadcasting of IP datagrams.
+
+CIPSO gateways MUST be capable of translating a CIPSO option from one
+DOI to another when forwarding datagrams between networks. For
+efficiency purposes this capability is only a desired feature for CIPSO
+routers.
+
+
+5.4 Label of ICMP Messages
+
+The CIPSO label to be used on all outgoing ICMP messages MUST be equivalent
+to the label of the datagram that caused the ICMP message. If the ICMP was
+generated due to a problem associated with the original CIPSO label then the
+following responses are allowed:
+
+ a. Use the CIPSO label of the original IP datagram
+ b. Drop the original datagram with no return message generated
+
+In most cases these options will have the same effect. If you can not
+interpret the label or if it is outside the label range of your host or
+interface then an ICMP message with the same label will probably not be
+able to exit the system.
+
+
+6. Assignment of DOI Identifier Numbers =
+
+Requests for assignment of a DOI identifier number should be addressed to
+the Internet Assigned Numbers Authority (IANA).
+
+
+7. Acknowledgements
+
+Much of the material in this RFC is based on (and copied from) work
+done by Gary Winiger of Sun Microsystems and published as Commercial
+IP Security Option at the INTEROP 89, Commercial IPSO Workshop.
+
+
+8. Author's Address
+
+To submit mail for distribution to members of the IETF CIPSO Working
+Group, send mail to: cipso@wdl1.wdl.loral.com.
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 11]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+
+To be added to or deleted from this distribution, send mail to:
+cipso-request@wdl1.wdl.loral.com.
+
+
+9. References
+
+RFC 1038, "Draft Revised IP Security Option", M. St. Johns, IETF, January
+1988.
+
+RFC 1108, "U.S. Department of Defense Security Options
+for the Internet Protocol", Stephen Kent, IAB, 1 March, 1991.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 12]
+
+
+
Index: linux-2.6.18-rc1/Documentation/netlabel/introduction.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/introduction.txt
@@ -0,0 +1,53 @@
+NetLabel Introduction
+==============================================================================
+Paul Moore, paul.moore@hp.com
+
+June 27, 2006
+
+ * Overview
+
+NetLabel is a mechanism which can be used by kernel security modules to attach
+security attributes to outgoing network packets generated from user space
+applications and read security attributes from incoming network packets. It
+is composed of four main components, the protocol engines, the communication
+layer, the kernel security module API, and the network stack operations.
+
+ * Protocol Engines
+
+The protocol engines are responsible for both applying and retrieving the
+network packet's security attributes. If any translation between the network
+security attributes and those on the host are required then the protocol
+engine will handle those tasks as well. Other kernel subsystems should
+refrain from calling the protocol engines directly, instead they should use
+the NetLabel kernel security module API described below.
+
+Detailed information about each NetLabel protocol engine can be found in this
+directory, consult '00-INDEX' for filenames.
+
+ * Communication Layer
+
+The communication layer exists to allow NetLabel configuration and monitoring
+from user space. The NetLabel communication layer uses a message based
+protocol built on top of the Generic NETLINK transport mechanism. The exact
+formatting of these NetLabel messages as well as the Generic NETLINK family
+names can be found in the the 'net/netlabel/' directory as comments in the
+header files as well as in 'include/net/netlabel.h'.
+
+ * Security Module API
+
+The purpose of the NetLabel security module API is to provide a protocol
+independent interface to the underlying NetLabel protocol engines. In addition
+to protocol independence, the security module API is designed to be completely
+LSM independent which should allow multiple LSMs to leverage the same code
+base.
+
+Detailed information about the NetLabel security module API can be found in the
+'include/net/netlabel.h' header file as well as the 'lsm_interface.txt' file
+found in this directory.
+
+ * Network Stack Operations
+
+The reason for the network stack operations are to ensure the correct handling
+of the security attributes for kernel generated packets. In general only the
+core network stack should call these functions.
+
Index: linux-2.6.18-rc1/Documentation/netlabel/lsm_interface.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/lsm_interface.txt
@@ -0,0 +1,47 @@
+NetLabel Linux Security Module Interface
+==============================================================================
+Paul Moore, paul.moore@hp.com
+
+May 17, 2006
+
+ * Overview
+
+NetLabel is a mechanism which can set and retrieve security attributes from
+network packets. It is intended to be used by LSM developers who want to make
+use of a common code base for several different packet labeling protocols.
+The NetLabel security module API is defined in 'include/net/netlabel.h' but a
+brief overview is given below.
+
+ * NetLabel Security Attributes
+
+Since NetLabel supports multiple different packet labeling protocols and LSMs
+it uses the concept of security attributes to refer to the packet's security
+labels. The NetLabel security attributes are defined by the
+'netlbl_lsm_secattr' structure in the NetLabel header file. Internally the
+NetLabel subsystem converts the security attributes to and from the correct
+low-level packet label depending on the NetLabel build time and run time
+configuration. It is up to the LSM developer to translate the NetLabel
+security attributes into whatever security identifiers are in use for their
+particular LSM.
+
+ * NetLabel LSM Protocol Operations
+
+These are the functions which allow the LSM developer to manipulate the labels
+on outgoing packets as well as read the labels on incoming packets. Functions
+exist to operate both on sockets as well as the sk_buffs directly. These high
+level functions are translated into low level protocol operations based on how
+the administrator has configured the NetLabel subsystem.
+
+ * NetLabel Label Mapping Cache Operations
+
+Depending on the exact configuration, translation between the network packet
+label and the internal LSM security identifier can be time consuming. The
+NetLabel label mapping cache is a caching mechanism which can be used to
+sidestep much of this overhead once a mapping has been established. Once the
+LSM has received a packet, used NetLabel to decode it's security attributes,
+and translated the security attributes into a LSM internal identifier the LSM
+can use the NetLabel caching functions to associate the LSM internal
+identifier with the network packet's label. This means that in the future
+when a incoming packet matches a cached value not only are the internal
+NetLabel translation mechanisms bypassed but the LSM translation mechanisms are
+bypassed as well which should result in a significant reduction in overhead.
Index: linux-2.6.18-rc1/Documentation/networking/ip-sysctl.txt
===================================================================
--- linux-2.6.18-rc1.orig/Documentation/networking/ip-sysctl.txt
+++ linux-2.6.18-rc1/Documentation/networking/ip-sysctl.txt
@@ -369,6 +369,34 @@ tcp_slow_start_after_idle - BOOLEAN
be timed out after an idle period.
Default: 1
+CIPSOv4 Variables:
+
+cipso_cache_enable - BOOLEAN
+ If set, enable additions to and lookups from the CIPSO label mapping
+ cache. If unset, additions are ignored and lookups always result in a
+ miss. However, regardless of the setting the cache is still
+ invalidated when required when means you can safely toggle this on and
+ off and the cache will always be "safe".
+
+cipso_cache_bucket_size - INTEGER
+ This is the maximum size of a CIPSO label cache bucket, once the limit
+ is reached the "older" entries inside the bucket are tossed to make
+ room for the new CIPSO labels.
+
+cipso_rbm_optfmt - BOOLEAN
+ Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
+ the CIPSO draft specification (see Documentation/netlabel for details).
+ This means that when set the CIPSO tag will be padded with empty
+ categories in order to make the packet data 32-bit aligned.
+
+cipso_rbm_structvalid - BOOLEAN
+ If set, do a very strict check of the CIPSO option when
+ ip_options_compile() is called. If unset, relax the checks done during
+ ip_options_compile(). Either way is "safe" as errors are caught else
+ where in the CIPSO processing code but setting this to 0 (False) should
+ result in less work (i.e. it should be faster) but possibile problems
+ with other implementations that require strict checks.
+
IP Variables:
ip_local_port_range - 2 INTEGERS
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 1/7] NetLabel: documentation
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Documentation for the NetLabel system, this includes a basic overview of how
NetLabel works, how LSM developers can integrate it into their favorite
LSM, as well as documentation on the CIPSO related sysctl variables. Also,
due to the difficulty of finding expired IETF drafts, I am including the IETF
CIPSO draft that is the basis of the NetLabel CIPSO implementation.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
CREDITS | 7
Documentation/00-INDEX | 2
Documentation/netlabel/00-INDEX | 10
Documentation/netlabel/cipso_ipv4.txt | 48
Documentation/netlabel/draft-ietf-cipso-ipsecurity-01.txt | 791 ++++++++++++++
Documentation/netlabel/introduction.txt | 53
Documentation/netlabel/lsm_interface.txt | 47
Documentation/networking/ip-sysctl.txt | 28
8 files changed, 986 insertions(+)
Index: linux-2.6.18-rc1/CREDITS
===================================================================
--- linux-2.6.18-rc1.orig/CREDITS
+++ linux-2.6.18-rc1/CREDITS
@@ -2384,6 +2384,13 @@ N: Thomas Molina
E: tmolina@cablespeed.com
D: bug fixes, documentation, minor hackery
+N: Paul Moore
+E: paul.moore@hp.com
+D: NetLabel author
+S: Hewlett-Packard
+S: 110 Spit Brook Road
+S: Nashua, NH 03062
+
N: James Morris
E: jmorris@namei.org
W: http://namei.org/
Index: linux-2.6.18-rc1/Documentation/00-INDEX
===================================================================
--- linux-2.6.18-rc1.orig/Documentation/00-INDEX
+++ linux-2.6.18-rc1/Documentation/00-INDEX
@@ -184,6 +184,8 @@ mtrr.txt
- how to use PPro Memory Type Range Registers to increase performance.
nbd.txt
- info on a TCP implementation of a network block device.
+netlabel/
+ - directory with information on the NetLabel subsystem.
networking/
- directory with info on various aspects of networking with Linux.
nfsroot.txt
Index: linux-2.6.18-rc1/Documentation/netlabel/00-INDEX
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/00-INDEX
@@ -0,0 +1,10 @@
+00-INDEX
+ - this file.
+cipso_ipv4.txt
+ - documentation on the IPv4 CIPSO protocol engine.
+draft-ietf-cipso-ipsecurity-01.txt
+ - IETF draft of the CIPSO protocol, dated 16 July 1992.
+introduction.txt
+ - NetLabel introduction, READ THIS FIRST.
+lsm_interface.txt
+ - documentation on the NetLabel kernel security module API.
Index: linux-2.6.18-rc1/Documentation/netlabel/cipso_ipv4.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/cipso_ipv4.txt
@@ -0,0 +1,48 @@
+NetLabel CIPSO/IPv4 Protocol Engine
+==============================================================================
+Paul Moore, paul.moore@hp.com
+
+May 17, 2006
+
+ * Overview
+
+The NetLabel CIPSO/IPv4 protocol engine is based on the IETF Commercial IP
+Security Option (CIPSO) draft from July 16, 1992. A copy of this draft can be
+found in this directory, consult '00-INDEX' for the filename. While the IETF
+draft never made it to an RFC standard it has become a de-facto standard for
+labeled networking and is used in many trusted operating systems.
+
+ * Outbound Packet Processing
+
+The CIPSO/IPv4 protocol engine applies the CIPSO IP option to packets by
+adding the CIPSO label to the socket. This causes all packets leaving the
+system through the socket to have the CIPSO IP option applied. The socket's
+CIPSO label can be changed at any point in time, however, it is recommended
+that it is set upon the socket's creation. The LSM can set the socket's CIPSO
+label by using the NetLabel security module API; if the NetLabel "domain" is
+configured to use CIPSO for packet labeling then a CIPSO IP option will be
+generated and attached to the socket.
+
+ * Inbound Packet Processing
+
+The CIPSO/IPv4 protocol engine validates every CIPSO IP option it finds at the
+IP layer without any special handling required by the LSM. However, in order
+to decode and translate the CIPSO label on the packet the LSM must use the
+NetLabel security module API to extract the security attributes of the packet.
+This is typically done at the socket layer using the 'socket_sock_rcv_skb()'
+LSM hook.
+
+ * Label Translation
+
+The CIPSO/IPv4 protocol engine contains a mechanism to translate CIPSO security
+attributes such as sensitivity level and category to values which are
+appropriate for the host. These mappings are defined as part of a CIPSO
+Domain Of Interpretation (DOI) definition and are configured through the
+NetLabel user space communication layer. Each DOI definition can have a
+different security attribute mapping table.
+
+ * Label Translation Cache
+
+The NetLabel system provides a framework for caching security attribute
+mappings from the network labels to the corresponding LSM identifiers. The
+CIPSO/IPv4 protocol engine supports this caching mechanism.
Index: linux-2.6.18-rc1/Documentation/netlabel/draft-ietf-cipso-ipsecurity-01.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/draft-ietf-cipso-ipsecurity-01.txt
@@ -0,0 +1,791 @@
+IETF CIPSO Working Group
+16 July, 1992
+
+
+
+ COMMERCIAL IP SECURITY OPTION (CIPSO 2.2)
+
+
+
+1. Status
+
+This Internet Draft provides the high level specification for a Commercial
+IP Security Option (CIPSO). This draft reflects the version as approved by
+the CIPSO IETF Working Group. Distribution of this memo is unlimited.
+
+This document is an Internet Draft. Internet Drafts are working documents
+of the Internet Engineering Task Force (IETF), its Areas, and its Working
+Groups. Note that other groups may also distribute working documents as
+Internet Drafts.
+
+Internet Drafts are draft documents valid for a maximum of six months.
+Internet Drafts may be updated, replaced, or obsoleted by other documents
+at any time. It is not appropriate to use Internet Drafts as reference
+material or to cite them other than as a "working draft" or "work in
+progress."
+
+Please check the I-D abstract listing contained in each Internet Draft
+directory to learn the current status of this or any other Internet Draft.
+
+
+
+
+2. Background
+
+Currently the Internet Protocol includes two security options. One of
+these options is the DoD Basic Security Option (BSO) (Type 130) which allows
+IP datagrams to be labeled with security classifications. This option
+provides sixteen security classifications and a variable number of handling
+restrictions. To handle additional security information, such as security
+categories or compartments, another security option (Type 133) exists and
+is referred to as the DoD Extended Security Option (ESO). The values for
+the fixed fields within these two options are administered by the Defense
+Information Systems Agency (DISA).
+
+Computer vendors are now building commercial operating systems with
+mandatory access controls and multi-level security. These systems are
+no longer built specifically for a particular group in the defense or
+intelligence communities. They are generally available commercial systems
+for use in a variety of government and civil sector environments.
+
+The small number of ESO format codes can not support all the possible
+applications of a commercial security option. The BSO and ESO were
+designed to only support the United States DoD. CIPSO has been designed
+to support multiple security policies. This Internet Draft provides the
+format and procedures required to support a Mandatory Access Control
+security policy. Support for additional security policies shall be
+defined in future RFCs.
+
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 1]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+
+3. CIPSO Format
+
+Option type: 134 (Class 0, Number 6, Copy on Fragmentation)
+Option length: Variable
+
+This option permits security related information to be passed between
+systems within a single Domain of Interpretation (DOI). A DOI is a
+collection of systems which agree on the meaning of particular values
+in the security option. An authority that has been assigned a DOI
+identifier will define a mapping between appropriate CIPSO field values
+and their human readable equivalent. This authority will distribute that
+mapping to hosts within the authority's domain. These mappings may be
+sensitive, therefore a DOI authority is not required to make these
+mappings available to anyone other than the systems that are included in
+the DOI.
+
+This option MUST be copied on fragmentation. This option appears at most
+once in a datagram. All multi-octet fields in the option are defined to be
+transmitted in network byte order. The format of this option is as follows:
+
++----------+----------+------//------+-----------//---------+
+| 10000110 | LLLLLLLL | DDDDDDDDDDDD | TTTTTTTTTTTTTTTTTTTT |
++----------+----------+------//------+-----------//---------+
+
+ TYPE=134 OPTION DOMAIN OF TAGS
+ LENGTH INTERPRETATION
+
+
+ Figure 1. CIPSO Format
+
+
+3.1 Type
+
+This field is 1 octet in length. Its value is 134.
+
+
+3.2 Length
+
+This field is 1 octet in length. It is the total length of the option
+including the type and length fields. With the current IP header length
+restriction of 40 octets the value of this field MUST not exceed 40.
+
+
+3.3 Domain of Interpretation Identifier
+
+This field is an unsigned 32 bit integer. The value 0 is reserved and MUST
+not appear as the DOI identifier in any CIPSO option. Implementations
+should assume that the DOI identifier field is not aligned on any particular
+byte boundary.
+
+To conserve space in the protocol, security levels and categories are
+represented by numbers rather than their ASCII equivalent. This requires
+a mapping table within CIPSO hosts to map these numbers to their
+corresponding ASCII representations. Non-related groups of systems may
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 2]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+have their own unique mappings. For example, one group of systems may
+use the number 5 to represent Unclassified while another group may use the
+number 1 to represent that same security level. The DOI identifier is used
+to identify which mapping was used for the values within the option.
+
+
+3.4 Tag Types
+
+A common format for passing security related information is necessary
+for interoperability. CIPSO uses sets of "tags" to contain the security
+information relevant to the data in the IP packet. Each tag begins with
+a tag type identifier followed by the length of the tag and ends with the
+actual security information to be passed. All multi-octet fields in a tag
+are defined to be transmitted in network byte order. Like the DOI
+identifier field in the CIPSO header, implementations should assume that
+all tags, as well as fields within a tag, are not aligned on any particular
+octet boundary. The tag types defined in this document contain alignment
+bytes to assist alignment of some information, however alignment can not
+be guaranteed if CIPSO is not the first IP option.
+
+CIPSO tag types 0 through 127 are reserved for defining standard tag
+formats. Their definitions will be published in RFCs. Tag types whose
+identifiers are greater than 127 are defined by the DOI authority and may
+only be meaningful in certain Domains of Interpretation. For these tag
+types, implementations will require the DOI identifier as well as the tag
+number to determine the security policy and the format associated with the
+tag. Use of tag types above 127 are restricted to closed networks where
+interoperability with other networks will not be an issue. Implementations
+that support a tag type greater than 127 MUST support at least one DOI that
+requires only tag types 1 to 127.
+
+Tag type 0 is reserved. Tag types 1, 2, and 5 are defined in this
+Internet Draft. Types 3 and 4 are reserved for work in progress.
+The standard format for all current and future CIPSO tags is shown below:
+
++----------+----------+--------//--------+
+| TTTTTTTT | LLLLLLLL | IIIIIIIIIIIIIIII |
++----------+----------+--------//--------+
+ TAG TAG TAG
+ TYPE LENGTH INFORMATION
+
+ Figure 2: Standard Tag Format
+
+In the three tag types described in this document, the length and count
+restrictions are based on the current IP limitation of 40 octets for all
+IP options. If the IP header is later expanded, then the length and count
+restrictions specified in this document may increase to use the full area
+provided for IP options.
+
+
+3.4.1 Tag Type Classes
+
+Tag classes consist of tag types that have common processing requirements
+and support the same security policy. The three tags defined in this
+Internet Draft belong to the Mandatory Access Control (MAC) Sensitivity
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 3]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+class and support the MAC Sensitivity security policy.
+
+
+3.4.2 Tag Type 1
+
+This is referred to as the "bit-mapped" tag type. Tag type 1 is included
+in the MAC Sensitivity tag type class. The format of this tag type is as
+follows:
+
++----------+----------+----------+----------+--------//---------+
+| 00000001 | LLLLLLLL | 00000000 | LLLLLLLL | CCCCCCCCCCCCCCCCC |
++----------+----------+----------+----------+--------//---------+
+
+ TAG TAG ALIGNMENT SENSITIVITY BIT MAP OF
+ TYPE LENGTH OCTET LEVEL CATEGORIES
+
+ Figure 3. Tag Type 1 Format
+
+
+3.4.2.1 Tag Type
+
+This field is 1 octet in length and has a value of 1.
+
+
+3.4.2.2 Tag Length
+
+This field is 1 octet in length. It is the total length of the tag type
+including the type and length fields. With the current IP header length
+restriction of 40 bytes the value within this field is between 4 and 34.
+
+
+3.4.2.3 Alignment Octet
+
+This field is 1 octet in length and always has the value of 0. Its purpose
+is to align the category bitmap field on an even octet boundary. This will
+speed many implementations including router implementations.
+
+
+3.4.2.4 Sensitivity Level
+
+This field is 1 octet in length. Its value is from 0 to 255. The values
+are ordered with 0 being the minimum value and 255 representing the maximum
+value.
+
+
+3.4.2.5 Bit Map of Categories
+
+The length of this field is variable and ranges from 0 to 30 octets. This
+provides representation of categories 0 to 239. The ordering of the bits
+is left to right or MSB to LSB. For example category 0 is represented by
+the most significant bit of the first byte and category 15 is represented
+by the least significant bit of the second byte. Figure 4 graphically
+shows this ordering. Bit N is binary 1 if category N is part of the label
+for the datagram, and bit N is binary 0 if category N is not part of the
+label. Except for the optimized tag 1 format described in the next section,
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 4]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+minimal encoding SHOULD be used resulting in no trailing zero octets in the
+category bitmap.
+
+ octet 0 octet 1 octet 2 octet 3 octet 4 octet 5
+ XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX . . .
+bit 01234567 89111111 11112222 22222233 33333333 44444444
+number 012345 67890123 45678901 23456789 01234567
+
+ Figure 4. Ordering of Bits in Tag 1 Bit Map
+
+
+3.4.2.6 Optimized Tag 1 Format
+
+Routers work most efficiently when processing fixed length fields. To
+support these routers there is an optimized form of tag type 1. The format
+does not change. The only change is to the category bitmap which is set to
+a constant length of 10 octets. Trailing octets required to fill out the 10
+octets are zero filled. Ten octets, allowing for 80 categories, was chosen
+because it makes the total length of the CIPSO option 20 octets. If CIPSO
+is the only option then the option will be full word aligned and additional
+filler octets will not be required.
+
+
+3.4.3 Tag Type 2
+
+This is referred to as the "enumerated" tag type. It is used to describe
+large but sparsely populated sets of categories. Tag type 2 is in the MAC
+Sensitivity tag type class. The format of this tag type is as follows:
+
++----------+----------+----------+----------+-------------//-------------+
+| 00000010 | LLLLLLLL | 00000000 | LLLLLLLL | CCCCCCCCCCCCCCCCCCCCCCCCCC |
++----------+----------+----------+----------+-------------//-------------+
+
+ TAG TAG ALIGNMENT SENSITIVITY ENUMERATED
+ TYPE LENGTH OCTET LEVEL CATEGORIES
+
+ Figure 5. Tag Type 2 Format
+
+
+3.4.3.1 Tag Type
+
+This field is one octet in length and has a value of 2.
+
+
+3.4.3.2 Tag Length
+
+This field is 1 octet in length. It is the total length of the tag type
+including the type and length fields. With the current IP header length
+restriction of 40 bytes the value within this field is between 4 and 34.
+
+
+3.4.3.3 Alignment Octet
+
+This field is 1 octet in length and always has the value of 0. Its purpose
+is to align the category field on an even octet boundary. This will
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 5]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+speed many implementations including router implementations.
+
+
+3.4.3.4 Sensitivity Level
+
+This field is 1 octet in length. Its value is from 0 to 255. The values
+are ordered with 0 being the minimum value and 255 representing the
+maximum value.
+
+
+3.4.3.5 Enumerated Categories
+
+In this tag, categories are represented by their actual value rather than
+by their position within a bit field. The length of each category is 2
+octets. Up to 15 categories may be represented by this tag. Valid values
+for categories are 0 to 65534. Category 65535 is not a valid category
+value. The categories MUST be listed in ascending order within the tag.
+
+
+3.4.4 Tag Type 5
+
+This is referred to as the "range" tag type. It is used to represent
+labels where all categories in a range, or set of ranges, are included
+in the sensitivity label. Tag type 5 is in the MAC Sensitivity tag type
+class. The format of this tag type is as follows:
+
++----------+----------+----------+----------+------------//-------------+
+| 00000101 | LLLLLLLL | 00000000 | LLLLLLLL | Top/Bottom | Top/Bottom |
++----------+----------+----------+----------+------------//-------------+
+
+ TAG TAG ALIGNMENT SENSITIVITY CATEGORY RANGES
+ TYPE LENGTH OCTET LEVEL
+
+ Figure 6. Tag Type 5 Format
+
+
+3.4.4.1 Tag Type
+
+This field is one octet in length and has a value of 5.
+
+
+3.4.4.2 Tag Length
+
+This field is 1 octet in length. It is the total length of the tag type
+including the type and length fields. With the current IP header length
+restriction of 40 bytes the value within this field is between 4 and 34.
+
+
+3.4.4.3 Alignment Octet
+
+This field is 1 octet in length and always has the value of 0. Its purpose
+is to align the category range field on an even octet boundary. This will
+speed many implementations including router implementations.
+
+
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 6]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+3.4.4.4 Sensitivity Level
+
+This field is 1 octet in length. Its value is from 0 to 255. The values
+are ordered with 0 being the minimum value and 255 representing the maximum
+value.
+
+
+3.4.4.5 Category Ranges
+
+A category range is a 4 octet field comprised of the 2 octet index of the
+highest numbered category followed by the 2 octet index of the lowest
+numbered category. These range endpoints are inclusive within the range of
+categories. All categories within a range are included in the sensitivity
+label. This tag may contain a maximum of 7 category pairs. The bottom
+category endpoint for the last pair in the tag MAY be omitted and SHOULD be
+assumed to be 0. The ranges MUST be non-overlapping and be listed in
+descending order. Valid values for categories are 0 to 65534. Category
+65535 is not a valid category value.
+
+
+3.4.5 Minimum Requirements
+
+A CIPSO implementation MUST be capable of generating at least tag type 1 in
+the non-optimized form. In addition, a CIPSO implementation MUST be able
+to receive any valid tag type 1 even those using the optimized tag type 1
+format.
+
+
+4. Configuration Parameters
+
+The configuration parameters defined below are required for all CIPSO hosts,
+gateways, and routers that support multiple sensitivity labels. A CIPSO
+host is defined to be the origination or destination system for an IP
+datagram. A CIPSO gateway provides IP routing services between two or more
+IP networks and may be required to perform label translations between
+networks. A CIPSO gateway may be an enhanced CIPSO host or it may just
+provide gateway services with no end system CIPSO capabilities. A CIPSO
+router is a dedicated IP router that routes IP datagrams between two or more
+IP networks.
+
+An implementation of CIPSO on a host MUST have the capability to reject a
+datagram for reasons that the information contained can not be adequately
+protected by the receiving host or if acceptance may result in violation of
+the host or network security policy. In addition, a CIPSO gateway or router
+MUST be able to reject datagrams going to networks that can not provide
+adequate protection or may violate the network's security policy. To
+provide this capability the following minimal set of configuration
+parameters are required for CIPSO implementations:
+
+HOST_LABEL_MAX - This parameter contains the maximum sensitivity label that
+a CIPSO host is authorized to handle. All datagrams that have a label
+greater than this maximum MUST be rejected by the CIPSO host. This
+parameter does not apply to CIPSO gateways or routers. This parameter need
+not be defined explicitly as it can be implicitly derived from the
+PORT_LABEL_MAX parameters for the associated interfaces.
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 7]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+
+HOST_LABEL_MIN - This parameter contains the minimum sensitivity label that
+a CIPSO host is authorized to handle. All datagrams that have a label less
+than this minimum MUST be rejected by the CIPSO host. This parameter does
+not apply to CIPSO gateways or routers. This parameter need not be defined
+explicitly as it can be implicitly derived from the PORT_LABEL_MIN
+parameters for the associated interfaces.
+
+PORT_LABEL_MAX - This parameter contains the maximum sensitivity label for
+all datagrams that may exit a particular network interface port. All
+outgoing datagrams that have a label greater than this maximum MUST be
+rejected by the CIPSO system. The label within this parameter MUST be
+less than or equal to the label within the HOST_LABEL_MAX parameter. This
+parameter does not apply to CIPSO hosts that support only one network port.
+
+PORT_LABEL_MIN - This parameter contains the minimum sensitivity label for
+all datagrams that may exit a particular network interface port. All
+outgoing datagrams that have a label less than this minimum MUST be
+rejected by the CIPSO system. The label within this parameter MUST be
+greater than or equal to the label within the HOST_LABEL_MIN parameter.
+This parameter does not apply to CIPSO hosts that support only one network
+port.
+
+PORT_DOI - This parameter is used to assign a DOI identifier value to a
+particular network interface port. All CIPSO labels within datagrams
+going out this port MUST use the specified DOI identifier. All CIPSO
+hosts and gateways MUST support either this parameter, the NET_DOI
+parameter, or the HOST_DOI parameter.
+
+NET_DOI - This parameter is used to assign a DOI identifier value to a
+particular IP network address. All CIPSO labels within datagrams destined
+for the particular IP network MUST use the specified DOI identifier. All
+CIPSO hosts and gateways MUST support either this parameter, the PORT_DOI
+parameter, or the HOST_DOI parameter.
+
+HOST_DOI - This parameter is used to assign a DOI identifier value to a
+particular IP host address. All CIPSO labels within datagrams destined for
+the particular IP host will use the specified DOI identifier. All CIPSO
+hosts and gateways MUST support either this parameter, the PORT_DOI
+parameter, or the NET_DOI parameter.
+
+This list represents the minimal set of configuration parameters required
+to be compliant. Implementors are encouraged to add to this list to
+provide enhanced functionality and control. For example, many security
+policies may require both incoming and outgoing datagrams be checked against
+the port and host label ranges.
+
+
+4.1 Port Range Parameters
+
+The labels represented by the PORT_LABEL_MAX and PORT_LABEL_MIN parameters
+MAY be in CIPSO or local format. Some CIPSO systems, such as routers, may
+want to have the range parameters expressed in CIPSO format so that incoming
+labels do not have to be converted to a local format before being compared
+against the range. If multiple DOIs are supported by one of these CIPSO
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 8]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+systems then multiple port range parameters would be needed, one set for
+each DOI supported on a particular port.
+
+The port range will usually represent the total set of labels that may
+exist on the logical network accessed through the corresponding network
+interface. It may, however, represent a subset of these labels that are
+allowed to enter the CIPSO system.
+
+
+4.2 Single Label CIPSO Hosts
+
+CIPSO implementations that support only one label are not required to
+support the parameters described above. These limited implementations are
+only required to support a NET_LABEL parameter. This parameter contains
+the CIPSO label that may be inserted in datagrams that exit the host. In
+addition, the host MUST reject any incoming datagram that has a label which
+is not equivalent to the NET_LABEL parameter.
+
+
+5. Handling Procedures
+
+This section describes the processing requirements for incoming and
+outgoing IP datagrams. Just providing the correct CIPSO label format
+is not enough. Assumptions will be made by one system on how a
+receiving system will handle the CIPSO label. Wrong assumptions may
+lead to non-interoperability or even a security incident. The
+requirements described below represent the minimal set needed for
+interoperability and that provide users some level of confidence.
+Many other requirements could be added to increase user confidence,
+however at the risk of restricting creativity and limiting vendor
+participation.
+
+
+5.1 Input Procedures
+
+All datagrams received through a network port MUST have a security label
+associated with them, either contained in the datagram or assigned to the
+receiving port. Without this label the host, gateway, or router will not
+have the information it needs to make security decisions. This security
+label will be obtained from the CIPSO if the option is present in the
+datagram. See section 4.1.2 for handling procedures for unlabeled
+datagrams. This label will be compared against the PORT (if appropriate)
+and HOST configuration parameters defined in section 3.
+
+If any field within the CIPSO option, such as the DOI identifier, is not
+recognized the IP datagram is discarded and an ICMP "parameter problem"
+(type 12) is generated and returned. The ICMP code field is set to "bad
+parameter" (code 0) and the pointer is set to the start of the CIPSO field
+that is unrecognized.
+
+If the contents of the CIPSO are valid but the security label is
+outside of the configured host or port label range, the datagram is
+discarded and an ICMP "destination unreachable" (type 3) is generated
+and returned. The code field of the ICMP is set to "communication with
+destination network administratively prohibited" (code 9) or to
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 9]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+"communication with destination host administratively prohibited"
+(code 10). The value of the code field used is dependent upon whether
+the originator of the ICMP message is acting as a CIPSO host or a CIPSO
+gateway. The recipient of the ICMP message MUST be able to handle either
+value. The same procedure is performed if a CIPSO can not be added to an
+IP packet because it is too large to fit in the IP options area.
+
+If the error is triggered by receipt of an ICMP message, the message
+is discarded and no response is permitted (consistent with general ICMP
+processing rules).
+
+
+5.1.1 Unrecognized tag types
+
+The default condition for any CIPSO implementation is that an
+unrecognized tag type MUST be treated as a "parameter problem" and
+handled as described in section 4.1. A CIPSO implementation MAY allow
+the system administrator to identify tag types that may safely be
+ignored. This capability is an allowable enhancement, not a
+requirement.
+
+
+5.1.2 Unlabeled Packets
+
+A network port may be configured to not require a CIPSO label for all
+incoming datagrams. For this configuration a CIPSO label must be
+assigned to that network port and associated with all unlabeled IP
+datagrams. This capability might be used for single level networks or
+networks that have CIPSO and non-CIPSO hosts and the non-CIPSO hosts
+all operate at the same label.
+
+If a CIPSO option is required and none is found, the datagram is
+discarded and an ICMP "parameter problem" (type 12) is generated and
+returned to the originator of the datagram. The code field of the ICMP
+is set to "option missing" (code 1) and the ICMP pointer is set to 134
+(the value of the option type for the missing CIPSO option).
+
+
+5.2 Output Procedures
+
+A CIPSO option MUST appear only once in a datagram. Only one tag type
+from the MAC Sensitivity class MAY be included in a CIPSO option. Given
+the current set of defined tag types, this means that CIPSO labels at
+first will contain only one tag.
+
+All datagrams leaving a CIPSO system MUST meet the following condition:
+
+ PORT_LABEL_MIN <= CIPSO label <= PORT_LABEL_MAX
+
+If this condition is not satisfied the datagram MUST be discarded.
+If the CIPSO system only supports one port, the HOST_LABEL_MIN and the
+HOST_LABEL_MAX parameters MAY be substituted for the PORT parameters in
+the above condition.
+
+The DOI identifier to be used for all outgoing datagrams is configured by
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 10]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+the administrator. If port level DOI identifier assignment is used, then
+the PORT_DOI configuration parameter MUST contain the DOI identifier to
+use. If network level DOI assignment is used, then the NET_DOI parameter
+MUST contain the DOI identifier to use. And if host level DOI assignment
+is employed, then the HOST_DOI parameter MUST contain the DOI identifier
+to use. A CIPSO implementation need only support one level of DOI
+assignment.
+
+
+5.3 DOI Processing Requirements
+
+A CIPSO implementation MUST support at least one DOI and SHOULD support
+multiple DOIs. System and network administrators are cautioned to
+ensure that at least one DOI is common within an IP network to allow for
+broadcasting of IP datagrams.
+
+CIPSO gateways MUST be capable of translating a CIPSO option from one
+DOI to another when forwarding datagrams between networks. For
+efficiency purposes this capability is only a desired feature for CIPSO
+routers.
+
+
+5.4 Label of ICMP Messages
+
+The CIPSO label to be used on all outgoing ICMP messages MUST be equivalent
+to the label of the datagram that caused the ICMP message. If the ICMP was
+generated due to a problem associated with the original CIPSO label then the
+following responses are allowed:
+
+ a. Use the CIPSO label of the original IP datagram
+ b. Drop the original datagram with no return message generated
+
+In most cases these options will have the same effect. If you can not
+interpret the label or if it is outside the label range of your host or
+interface then an ICMP message with the same label will probably not be
+able to exit the system.
+
+
+6. Assignment of DOI Identifier Numbers =
+
+Requests for assignment of a DOI identifier number should be addressed to
+the Internet Assigned Numbers Authority (IANA).
+
+
+7. Acknowledgements
+
+Much of the material in this RFC is based on (and copied from) work
+done by Gary Winiger of Sun Microsystems and published as Commercial
+IP Security Option at the INTEROP 89, Commercial IPSO Workshop.
+
+
+8. Author's Address
+
+To submit mail for distribution to members of the IETF CIPSO Working
+Group, send mail to: cipso@wdl1.wdl.loral.com.
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 11]
+
+
+
+CIPSO INTERNET DRAFT 16 July, 1992
+
+
+
+
+To be added to or deleted from this distribution, send mail to:
+cipso-request@wdl1.wdl.loral.com.
+
+
+9. References
+
+RFC 1038, "Draft Revised IP Security Option", M. St. Johns, IETF, January
+1988.
+
+RFC 1108, "U.S. Department of Defense Security Options
+for the Internet Protocol", Stephen Kent, IAB, 1 March, 1991.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+Internet Draft, Expires 15 Jan 93 [PAGE 12]
+
+
+
Index: linux-2.6.18-rc1/Documentation/netlabel/introduction.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/introduction.txt
@@ -0,0 +1,53 @@
+NetLabel Introduction
+==============================================================================
+Paul Moore, paul.moore@hp.com
+
+June 27, 2006
+
+ * Overview
+
+NetLabel is a mechanism which can be used by kernel security modules to attach
+security attributes to outgoing network packets generated from user space
+applications and read security attributes from incoming network packets. It
+is composed of four main components, the protocol engines, the communication
+layer, the kernel security module API, and the network stack operations.
+
+ * Protocol Engines
+
+The protocol engines are responsible for both applying and retrieving the
+network packet's security attributes. If any translation between the network
+security attributes and those on the host are required then the protocol
+engine will handle those tasks as well. Other kernel subsystems should
+refrain from calling the protocol engines directly, instead they should use
+the NetLabel kernel security module API described below.
+
+Detailed information about each NetLabel protocol engine can be found in this
+directory, consult '00-INDEX' for filenames.
+
+ * Communication Layer
+
+The communication layer exists to allow NetLabel configuration and monitoring
+from user space. The NetLabel communication layer uses a message based
+protocol built on top of the Generic NETLINK transport mechanism. The exact
+formatting of these NetLabel messages as well as the Generic NETLINK family
+names can be found in the the 'net/netlabel/' directory as comments in the
+header files as well as in 'include/net/netlabel.h'.
+
+ * Security Module API
+
+The purpose of the NetLabel security module API is to provide a protocol
+independent interface to the underlying NetLabel protocol engines. In addition
+to protocol independence, the security module API is designed to be completely
+LSM independent which should allow multiple LSMs to leverage the same code
+base.
+
+Detailed information about the NetLabel security module API can be found in the
+'include/net/netlabel.h' header file as well as the 'lsm_interface.txt' file
+found in this directory.
+
+ * Network Stack Operations
+
+The reason for the network stack operations are to ensure the correct handling
+of the security attributes for kernel generated packets. In general only the
+core network stack should call these functions.
+
Index: linux-2.6.18-rc1/Documentation/netlabel/lsm_interface.txt
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/Documentation/netlabel/lsm_interface.txt
@@ -0,0 +1,47 @@
+NetLabel Linux Security Module Interface
+==============================================================================
+Paul Moore, paul.moore@hp.com
+
+May 17, 2006
+
+ * Overview
+
+NetLabel is a mechanism which can set and retrieve security attributes from
+network packets. It is intended to be used by LSM developers who want to make
+use of a common code base for several different packet labeling protocols.
+The NetLabel security module API is defined in 'include/net/netlabel.h' but a
+brief overview is given below.
+
+ * NetLabel Security Attributes
+
+Since NetLabel supports multiple different packet labeling protocols and LSMs
+it uses the concept of security attributes to refer to the packet's security
+labels. The NetLabel security attributes are defined by the
+'netlbl_lsm_secattr' structure in the NetLabel header file. Internally the
+NetLabel subsystem converts the security attributes to and from the correct
+low-level packet label depending on the NetLabel build time and run time
+configuration. It is up to the LSM developer to translate the NetLabel
+security attributes into whatever security identifiers are in use for their
+particular LSM.
+
+ * NetLabel LSM Protocol Operations
+
+These are the functions which allow the LSM developer to manipulate the labels
+on outgoing packets as well as read the labels on incoming packets. Functions
+exist to operate both on sockets as well as the sk_buffs directly. These high
+level functions are translated into low level protocol operations based on how
+the administrator has configured the NetLabel subsystem.
+
+ * NetLabel Label Mapping Cache Operations
+
+Depending on the exact configuration, translation between the network packet
+label and the internal LSM security identifier can be time consuming. The
+NetLabel label mapping cache is a caching mechanism which can be used to
+sidestep much of this overhead once a mapping has been established. Once the
+LSM has received a packet, used NetLabel to decode it's security attributes,
+and translated the security attributes into a LSM internal identifier the LSM
+can use the NetLabel caching functions to associate the LSM internal
+identifier with the network packet's label. This means that in the future
+when a incoming packet matches a cached value not only are the internal
+NetLabel translation mechanisms bypassed but the LSM translation mechanisms are
+bypassed as well which should result in a significant reduction in overhead.
Index: linux-2.6.18-rc1/Documentation/networking/ip-sysctl.txt
===================================================================
--- linux-2.6.18-rc1.orig/Documentation/networking/ip-sysctl.txt
+++ linux-2.6.18-rc1/Documentation/networking/ip-sysctl.txt
@@ -369,6 +369,34 @@ tcp_slow_start_after_idle - BOOLEAN
be timed out after an idle period.
Default: 1
+CIPSOv4 Variables:
+
+cipso_cache_enable - BOOLEAN
+ If set, enable additions to and lookups from the CIPSO label mapping
+ cache. If unset, additions are ignored and lookups always result in a
+ miss. However, regardless of the setting the cache is still
+ invalidated when required when means you can safely toggle this on and
+ off and the cache will always be "safe".
+
+cipso_cache_bucket_size - INTEGER
+ This is the maximum size of a CIPSO label cache bucket, once the limit
+ is reached the "older" entries inside the bucket are tossed to make
+ room for the new CIPSO labels.
+
+cipso_rbm_optfmt - BOOLEAN
+ Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
+ the CIPSO draft specification (see Documentation/netlabel for details).
+ This means that when set the CIPSO tag will be padded with empty
+ categories in order to make the packet data 32-bit aligned.
+
+cipso_rbm_structvalid - BOOLEAN
+ If set, do a very strict check of the CIPSO option when
+ ip_options_compile() is called. If unset, relax the checks done during
+ ip_options_compile(). Either way is "safe" as errors are caught else
+ where in the CIPSO processing code but setting this to 0 (False) should
+ result in less work (i.e. it should be faster) but possibile problems
+ with other implementations that require strict checks.
+
IP Variables:
ip_local_port_range - 2 INTEGERS
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 2/7] NetLabel: core network changes
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-net_core-2.6.18 --]
[-- Type: text/plain, Size: 28543 bytes --]
Changes to the core network stack to support the NetLabel subsystem. This
includes changes to the IPv4 option handling to support CIPSO labels, and a new
NetLabel hook in inet_accept() to handle NetLabel attributes across accept()
calls done by in-kernel daemons.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
include/linux/ip.h | 1
include/net/cipso_ipv4.h | 266 ++++++++++++++++++++
include/net/inet_sock.h | 2
include/net/netlabel.h | 615 +++++++++++++++++++++++++++++++++++++++++++++++
net/ipv4/af_inet.c | 3
net/ipv4/ah4.c | 2
net/ipv4/ip_options.c | 19 +
7 files changed, 906 insertions(+), 2 deletions(-)
Index: linux-2.6.18-rc1/include/linux/ip.h
===================================================================
--- linux-2.6.18-rc1.orig/include/linux/ip.h
+++ linux-2.6.18-rc1/include/linux/ip.h
@@ -57,6 +57,7 @@
#define IPOPT_SEC (2 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_LSRR (3 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_TIMESTAMP (4 |IPOPT_MEASUREMENT)
+#define IPOPT_CIPSO (6 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_RR (7 |IPOPT_CONTROL)
#define IPOPT_SID (8 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_SSRR (9 |IPOPT_CONTROL|IPOPT_COPY)
Index: linux-2.6.18-rc1/include/net/cipso_ipv4.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/include/net/cipso_ipv4.h
@@ -0,0 +1,266 @@
+/*
+ * CIPSO - Commercial IP Security Option
+ *
+ * This is an implementation of the CIPSO 2.2 protocol as specified in
+ * draft-ietf-cipso-ipsecurity-01.txt with additional tag types as found in
+ * FIPS-188, copies of both documents can be found in the Documentation
+ * directory. While CIPSO never became a full IETF RFC standard many vendors
+ * have chosen to adopt the protocol and over the years it has become a
+ * de-facto standard for labeled networking.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _CIPSO_IPV4_H
+#define _CIPSO_IPV4_H
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <net/netlabel.h>
+
+/* known doi values */
+#define CIPSO_V4_DOI_UNKNOWN 0x00000000
+
+/* tag types */
+#define CIPSO_V4_TAG_INVALID 0
+#define CIPSO_V4_TAG_RBITMAP 1
+#define CIPSO_V4_TAG_ENUM 2
+#define CIPSO_V4_TAG_RANGE 5
+#define CIPSO_V4_TAG_PBITMAP 6
+#define CIPSO_V4_TAG_FREEFORM 7
+
+/* doi mapping types */
+#define CIPSO_V4_MAP_UNKNOWN 0
+#define CIPSO_V4_MAP_STD 1
+#define CIPSO_V4_MAP_PASS 2
+
+/* limits */
+#define CIPSO_V4_MAX_REM_LVLS 256
+#define CIPSO_V4_INV_LVL 0x80000000
+#define CIPSO_V4_MAX_LOC_LVLS (CIPSO_V4_INV_LVL - 1)
+#define CIPSO_V4_MAX_REM_CATS 65536
+#define CIPSO_V4_INV_CAT 0x80000000
+#define CIPSO_V4_MAX_LOC_CATS (CIPSO_V4_INV_CAT - 1)
+
+/*
+ * CIPSO DOI definitions
+ */
+
+/* DOI definition struct */
+#define CIPSO_V4_TAG_MAXCNT 5
+struct cipso_v4_doi {
+ u32 doi;
+ u32 type;
+ union {
+ struct cipso_v4_std_map_tbl *std;
+ } map;
+ u8 tags[CIPSO_V4_TAG_MAXCNT];
+
+ u32 valid;
+ struct list_head list;
+ struct rcu_head rcu;
+ struct list_head dom_list;
+};
+
+/* Standard CIPSO mapping table */
+/* NOTE: the highest order bit (i.e. 0x80000000) is an 'invalid' flag, if the
+ * bit is set then consider that value as unspecified, meaning the
+ * mapping for that particular level/category is invalid */
+struct cipso_v4_std_map_tbl {
+ struct {
+ u32 *cipso;
+ u32 *local;
+ u32 cipso_size;
+ u32 local_size;
+ } lvl;
+ struct {
+ u32 *cipso;
+ u32 *local;
+ u32 cipso_size;
+ u32 local_size;
+ } cat;
+};
+
+/*
+ * Sysctl Variables
+ */
+
+#ifdef CONFIG_NETLABEL
+extern int cipso_v4_cache_enabled;
+extern int cipso_v4_cache_bucketsize;
+extern int cipso_v4_rbm_optfmt;
+extern int cipso_v4_rbm_strictvalid;
+#endif
+
+/*
+ * Helper Functions
+ */
+
+#define CIPSO_V4_OPTEXIST(x) (IPCB(x)->opt.cipso != 0)
+#define CIPSO_V4_OPTPTR(x) ((x)->nh.raw + IPCB(x)->opt.cipso)
+
+/*
+ * DOI List Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+int cipso_v4_doi_add(struct cipso_v4_doi *doi_def);
+int cipso_v4_doi_remove(const u32 doi,
+ void (*callback) (struct rcu_head * head));
+struct cipso_v4_doi *cipso_v4_doi_getdef(const u32 doi);
+struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom);
+struct sk_buff *cipso_v4_doi_dump(const u32 doi, const size_t headroom);
+int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def, const char *domain);
+int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+ const char *domain);
+#else
+static inline int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_doi_remove(const u32 doi,
+ void (*callback) (struct rcu_head * head))
+{
+ return 0;
+}
+
+static inline struct cipso_v4_doi *cipso_v4_doi_getdef(const u32 doi)
+{
+ return NULL;
+}
+
+static inline struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
+{
+ return NULL;
+}
+
+static inline struct sk_buff *cipso_v4_doi_dump(const u32 doi,
+ const size_t headroom)
+{
+ return NULL;
+}
+
+static inline int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def,
+ const char *domain)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+ const char *domain)
+{
+ return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Label Mapping Cache Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+void cipso_v4_cache_invalidate(void);
+int cipso_v4_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr);
+#else
+static inline void cipso_v4_cache_invalidate(void)
+{
+ return;
+}
+
+static inline int cipso_v4_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Protocol Handling Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+void cipso_v4_error(struct sk_buff *skb, const int error, const u32 gateway);
+int cipso_v4_socket_setopt(struct socket *sock,
+ unsigned char *opt,
+ u32 opt_len);
+int cipso_v4_socket_setattr(const struct socket *sock,
+ const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr);
+int cipso_v4_socket_getopt(const struct socket *sock,
+ unsigned char **opt,
+ u32 *opt_len);
+int cipso_v4_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr);
+int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr);
+int cipso_v4_validate(unsigned char **option);
+#else
+static inline void cipso_v4_error(struct sk_buff *skb,
+ const int error,
+ const u32 gateway)
+{
+ return;
+}
+
+static inline int cipso_v4_socket_setopt(struct socket *sock,
+ unsigned char *opt,
+ u32 opt_len)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_setattr(const struct socket *sock,
+ const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_getopt(const struct socket *sock,
+ unsigned char **opt,
+ u32 *opt_len)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_validate(unsigned char **option)
+{
+ return -ENOSYS;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif /* _CIPSO_IPV4_H */
Index: linux-2.6.18-rc1/include/net/inet_sock.h
===================================================================
--- linux-2.6.18-rc1.orig/include/net/inet_sock.h
+++ linux-2.6.18-rc1/include/net/inet_sock.h
@@ -51,7 +51,7 @@ struct ip_options {
ts_needtime:1,
ts_needaddr:1;
unsigned char router_alert;
- unsigned char __pad1;
+ unsigned char cipso;
unsigned char __pad2;
unsigned char __data[0];
};
Index: linux-2.6.18-rc1/include/net/netlabel.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/include/net/netlabel.h
@@ -0,0 +1,615 @@
+/*
+ * NetLabel System
+ *
+ * The NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_H
+#define _NETLABEL_H
+
+#include <linux/types.h>
+#include <linux/skbuff.h>
+#include <net/netlink.h>
+
+/*
+ * NetLabel - A management interface for maintaining network packet label
+ * mapping tables for explicit packet labling protocols.
+ *
+ * Network protocols such as CIPSO and RIPSO require a label translation layer
+ * to convert the label on the packet into something meaningful on the host
+ * machine. In the current Linux implementation these mapping tables live
+ * inside the kernel; NetLabel provides a mechanism for user space applications
+ * to manage these mapping tables.
+ *
+ * NetLabel makes use of the Generic NETLINK mechanism as a transport layer to
+ * send messages between kernel and user space. The general format of a
+ * NetLabel message is shown below:
+ *
+ * +-----------------+-------------------+--------- --- -- -
+ * | struct nlmsghdr | struct genlmsghdr | payload
+ * +-----------------+-------------------+--------- --- -- -
+ *
+ * The 'nlmsghdr' and 'genlmsghdr' structs should be dealt with like normal.
+ * The payload is dependent on the subsystem specified in the
+ * 'nlmsghdr->nlmsg_type' and should be defined below, supporting functions
+ * should be defined in the corresponding net/netlabel/netlabel_<subsys>.h|c
+ * file. All of the fields in the NetLabel payload should be aligned using
+ * the alignment functions provided.
+ *
+ */
+
+/*
+ * NetLabel NETLINK protocol
+ */
+
+#define NETLBL_PROTO_VERSION 1
+
+/* NetLabel NETLINK types/families */
+#define NETLBL_NLTYPE_NONE 0
+#define NETLBL_NLTYPE_MGMT 1
+#define NETLBL_NLTYPE_MGMT_NAME "NLBL_MGMT"
+#define NETLBL_NLTYPE_RIPSO 2
+#define NETLBL_NLTYPE_RIPSO_NAME "NLBL_RIPSO"
+#define NETLBL_NLTYPE_CIPSOV4 3
+#define NETLBL_NLTYPE_CIPSOV4_NAME "NLBL_CIPSOv4"
+#define NETLBL_NLTYPE_CIPSOV6 4
+#define NETLBL_NLTYPE_CIPSOV6_NAME "NLBL_CIPSOv6"
+#define NETLBL_NLTYPE_UNLABELED 5
+#define NETLBL_NLTYPE_UNLABELED_NAME "NLBL_UNLBL"
+
+/* NetLabel return codes */
+#define NETLBL_E_OK 0
+
+/*
+ * Helper functions
+ */
+
+#define NETLBL_LEN_U8 netlbl_align(1)
+#define NETLBL_LEN_U16 netlbl_align(2)
+#define NETLBL_LEN_U32 netlbl_align(4)
+/**
+ * netlbl_align - Align a NetLabel data chunk
+ * @len: the data chunk length
+ *
+ * Description:
+ * Return the aligned data chunk length.
+ *
+ */
+static inline size_t netlbl_align(const size_t length)
+{
+ return NLMSG_ALIGN(length);
+}
+
+/**
+ * netlbl_put_u8 - Write a u8 value into a buffer
+ * @buffer: the buffer
+ * @val: the value
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer.
+ *
+ */
+static inline void netlbl_put_u8(unsigned char *buffer, const u8 val)
+{
+ *(u8 *)buffer = val;
+}
+
+/**
+ * netlbl_put_u16 - Write a u16 value into a buffer
+ * @buffer: the buffer
+ * @val: the value
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer.
+ *
+ */
+static inline void netlbl_put_u16(unsigned char *buffer, const u16 val)
+{
+ *(u16 *)buffer = val;
+}
+
+/**
+ * netlbl_put_u32 - Write a u32 value into a buffer
+ * @buffer: the buffer
+ * @val: the value
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer.
+ *
+ */
+static inline void netlbl_put_u32(unsigned char *buffer, const u32 val)
+{
+ *(u32 *)buffer = val;
+}
+
+/**
+ * netlbl_put_hdr - Write a NETLINK header into a buffer
+ * @buffer: the buffer
+ * @msg_type: the NETLINK message type
+ * @msg_len: the NETLINK message length
+ * @msg_flags: the NETLINK message flags
+ * @msg_pid: the NETLINK message PID
+ * @msg_seq: the NETLINK message sequence number
+ *
+ * Description:
+ * Use the given values to write a NETLINK header into the given buffer.
+ *
+ */
+static inline void netlbl_put_hdr(unsigned char *buffer,
+ const u32 msg_type,
+ const u16 msg_len,
+ const u16 msg_flags,
+ const u32 msg_pid,
+ const u32 msg_seq)
+{
+ struct nlmsghdr *hdr = (struct nlmsghdr *)buffer;
+ hdr->nlmsg_len = msg_len;
+ hdr->nlmsg_type = msg_type;
+ hdr->nlmsg_flags = msg_flags;
+ hdr->nlmsg_seq = msg_seq;
+ hdr->nlmsg_pid = msg_pid;
+}
+
+/**
+ * netlbl_putinc_u8 - Write a u8 value into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_u8(unsigned char **buffer,
+ const u8 val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u8));
+ netlbl_put_u8(*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_putinc_u16 - Write a u16 value into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_u16(unsigned char **buffer,
+ const u16 val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u16));
+ netlbl_put_u16(*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_putinc_u32 - Write a u32 value into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_u32(unsigned char **buffer,
+ const u32 val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u32));
+ netlbl_put_u32(*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_putinc_str - Write a string into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the string specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_str(unsigned char **buffer,
+ const char *val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(strlen(val) + 1);
+ strcpy((char *)*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_put_hdr - Write a NETLINK header into a buffer and increment the ptr
+ * @buffer: the buffer
+ * @msg_type: the NETLINK message type
+ * @msg_len: the NETLINK message length
+ * @msg_flags: the NETLINK message flags
+ * @msg_pid: the NETLINK message PID
+ * @msg_seq: the NETLINK message sequence number
+ *
+ * Description:
+ * Use the given values to write a NETLINK header into the given buffer and
+ * then increment the buffer pointer past the header.
+ *
+ */
+static inline void netlbl_putinc_hdr(unsigned char **buffer,
+ const u32 msg_type,
+ const u16 msg_len,
+ const u16 msg_flags,
+ const u32 msg_pid,
+ const u32 msg_seq)
+{
+ netlbl_put_hdr(*buffer,
+ msg_type,
+ msg_len,
+ msg_flags,
+ msg_pid,
+ msg_seq);
+ *buffer += NLMSG_HDRLEN;
+}
+
+/**
+ * netlbl_get_u8 - Read a u8 value from a buffer
+ * @buffer: the buffer
+ *
+ * Description:
+ * Return a u8 value pointed to by @buffer.
+ *
+ */
+static inline u8 netlbl_get_u8(const unsigned char *buffer)
+{
+ return *(u8 *)buffer;
+}
+
+/**
+ * netlbl_get_u16 - Read a u16 value from a buffer
+ * @buffer: the buffer
+ *
+ * Description:
+ * Return a u16 value pointed to by @buffer.
+ *
+ */
+static inline u16 netlbl_get_u16(const unsigned char *buffer)
+{
+ return *(u16 *)buffer;
+}
+
+/**
+ * netlbl_get_u32 - Read a u32 value from a buffer
+ * @buffer: the buffer
+ *
+ * Description:
+ * Return a u32 value pointed to by @buffer.
+ *
+ */
+static inline u32 netlbl_get_u32(const unsigned char *buffer)
+{
+ return *(u32 *)buffer;
+}
+
+/**
+ * netlbl_getinc_u8 - Read a u8 value from a buffer and increment the buffer
+ * @buffer: the buffer
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Return a u8 value pointed to by @buffer and increment the buffer pointer
+ * past the value. If @rem_len is not NULL, decrement it by the field size.
+ *
+ */
+static inline u8 netlbl_getinc_u8(unsigned char **buffer, ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u8));
+ u8 val = netlbl_get_u8(*buffer);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+ return val;
+}
+
+/**
+ * netlbl_getinc_u16 - Read a u16 value from a buffer and increment the buffer
+ * @buffer: the buffer
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Return a u16 value pointed to by @buffer and increment the buffer pointer
+ * past the value. If @rem_len is not NULL, decrement it by the field size.
+ *
+ */
+static inline u16 netlbl_getinc_u16(unsigned char **buffer, ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u16));
+ u16 val = netlbl_get_u16(*buffer);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+ return val;
+}
+
+/**
+ * netlbl_getinc_u32 - Read a u32 value from a buffer and increment the buffer
+ * @buffer: the buffer
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Return a u32 value pointed to by @buffer and increment the buffer pointer
+ * past the value. If @rem_len is not NULL, decrement it by the field size.
+ *
+ */
+static inline u32 netlbl_getinc_u32(unsigned char **buffer, ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u32));
+ u32 val = netlbl_get_u32(*buffer);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+ return val;
+}
+
+/**
+ * netlbl_netlink_alloc_skb - Allocate a NETLINK message buffer
+ * @head: the amount of headroom in bytes
+ * @body: the desired size (minus headroom) in bytes
+ * @gfp_flags: the alloc flags to pass to alloc_skb()
+ *
+ * Description:
+ * Allocate a NETLINK message buffer based on the sizes given in @head and
+ * @body. If @head is greater than zero skb_reserve() is called to reserve
+ * @head bytes at the start of the buffer. Returns a valid sk_buff pointer on
+ * success, NULL on failure.
+ *
+ */
+static inline struct sk_buff *netlbl_netlink_alloc_skb(const size_t head,
+ const size_t body,
+ int gfp_flags)
+{
+ struct sk_buff *skb;
+
+ skb = alloc_skb(NLMSG_ALIGN(head + body), gfp_flags);
+ if (skb == NULL)
+ return NULL;
+ if (head > 0) {
+ skb_reserve(skb, head);
+ if (skb_tailroom(skb) < body) {
+ kfree_skb(skb);
+ return NULL;
+ }
+ }
+
+ return skb;
+}
+
+/*
+ * NetLabel - Kernel API for accessing the network packet label mappings.
+ *
+ * The following functions are provided for use by other kernel modules,
+ * specifically kernel LSM modules, to provide a consistent, transparent API
+ * for dealing with explicit packet labeling protocols such as CIPSO and
+ * RIPSO. The functions defined here are implemented in the
+ * net/netlabel/netlabel_kapi.c file.
+ *
+ */
+
+/* Domain mapping definition struct */
+struct netlbl_dom_map;
+
+/* Domain mapping operations */
+int netlbl_domhsh_remove(const char *domain);
+
+/* LSM security attributes */
+struct netlbl_lsm_cache {
+ void (*free) (const void *data);
+ void *data;
+};
+struct netlbl_lsm_secattr {
+ char *domain;
+
+ u32 mls_lvl;
+ u32 mls_lvl_vld;
+ unsigned char *mls_cat;
+ size_t mls_cat_len;
+
+ struct netlbl_lsm_cache cache;
+};
+
+/*
+ * LSM security attribute operations
+ */
+
+
+/**
+ * netlbl_secattr_init - Initialize a netlbl_lsm_secattr struct
+ * @secattr: the struct to initialize
+ *
+ * Description:
+ * Initialize an already allocated netlbl_lsm_secattr struct. Returns zero on
+ * success, negative values on error.
+ *
+ */
+static inline int netlbl_secattr_init(struct netlbl_lsm_secattr *secattr)
+{
+ memset(secattr, 0, sizeof(*secattr));
+ return 0;
+}
+
+/**
+ * netlbl_secattr_destroy - Clears a netlbl_lsm_secattr struct
+ * @secattr: the struct to clear
+ * @clear_cache: cache clear flag
+ *
+ * Description:
+ * Destroys the @secattr struct, including freeing all of the internal buffers.
+ * If @clear_cache is true then free the cache fields, otherwise leave them
+ * intact. The struct must be reset with a call to netlbl_secattr_init()
+ * before reuse.
+ *
+ */
+static inline void netlbl_secattr_destroy(struct netlbl_lsm_secattr *secattr,
+ const u32 clear_cache)
+{
+ if (clear_cache && secattr->cache.data != NULL && secattr->cache.free)
+ secattr->cache.free(secattr->cache.data);
+ kfree(secattr->domain);
+ kfree(secattr->mls_cat);
+}
+
+/**
+ * netlbl_secattr_alloc - Allocate and initialize a netlbl_lsm_secattr struct
+ * @flags: the memory allocation flags
+ *
+ * Description:
+ * Allocate and initialize a netlbl_lsm_secattr struct. Returns a valid
+ * pointer on success, or NULL on failure.
+ *
+ */
+static inline struct netlbl_lsm_secattr *netlbl_secattr_alloc(const int flags)
+{
+ return kzalloc(sizeof(struct netlbl_lsm_secattr), flags);
+}
+
+/**
+ * netlbl_secattr_free - Frees a netlbl_lsm_secattr struct
+ * @secattr: the struct to free
+ * @clear_cache: cache clear flag
+ *
+ * Description:
+ * Frees @secattr including all of the internal buffers. If @clear_cache is
+ * true then free the cache fields, otherwise leave them intact.
+ *
+ */
+static inline void netlbl_secattr_free(struct netlbl_lsm_secattr *secattr,
+ const u32 clear_cache)
+{
+ netlbl_secattr_destroy(secattr, clear_cache);
+ kfree(secattr);
+}
+
+/*
+ * LSM protocol operations
+ */
+
+#ifdef CONFIG_NETLABEL
+int netlbl_socket_setattr(const struct socket *sock,
+ const struct netlbl_lsm_secattr *secattr);
+int netlbl_socket_peekattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr);
+int netlbl_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr);
+int netlbl_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr);
+void netlbl_skbuff_err(struct sk_buff *skb, int error);
+#else
+static inline int netlbl_socket_setattr(const struct socket *sock,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int netlbl_socket_peekattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int netlbl_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int netlbl_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline void netlbl_skbuff_err(struct sk_buff *skb, int error)
+{
+ return;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * LSM label mapping cache operations
+ */
+
+#ifdef CONFIG_NETLABEL
+void netlbl_cache_invalidate(void);
+int netlbl_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr);
+#else
+static inline void netlbl_cache_invalidate(void)
+{
+ return;
+}
+
+static inline int netlbl_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Network stack operations
+ */
+
+#ifdef CONFIG_NETLABEL
+void netlbl_socket_inet_accept(struct socket *sock, struct socket *newsock);
+#else
+static inline void netlbl_socket_inet_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ return;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif /* _NETLABEL_H */
Index: linux-2.6.18-rc1/net/ipv4/af_inet.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/af_inet.c
+++ linux-2.6.18-rc1/net/ipv4/af_inet.c
@@ -115,6 +115,7 @@
#ifdef CONFIG_IP_MROUTE
#include <linux/mroute.h>
#endif
+#include <net/netlabel.h>
DEFINE_SNMP_STAT(struct linux_mib, net_statistics) __read_mostly;
@@ -617,6 +618,8 @@ int inet_accept(struct socket *sock, str
sock_graft(sk2, newsock);
+ netlbl_socket_inet_accept(sock, newsock);
+
newsock->state = SS_CONNECTED;
err = 0;
release_sock(sk2);
Index: linux-2.6.18-rc1/net/ipv4/ah4.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/ah4.c
+++ linux-2.6.18-rc1/net/ipv4/ah4.c
@@ -34,7 +34,7 @@ static int ip_clear_mutable_options(stru
switch (*optptr) {
case IPOPT_SEC:
case 0x85: /* Some "Extended Security" crap. */
- case 0x86: /* Another "Commercial Security" crap. */
+ case IPOPT_CIPSO:
case IPOPT_RA:
case 0x80|21: /* RFC1770 */
break;
Index: linux-2.6.18-rc1/net/ipv4/ip_options.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/ip_options.c
+++ linux-2.6.18-rc1/net/ipv4/ip_options.c
@@ -24,6 +24,7 @@
#include <net/ip.h>
#include <net/icmp.h>
#include <net/route.h>
+#include <net/cipso_ipv4.h>
/*
* Write options to IP header, record destination address to
@@ -194,6 +195,13 @@ int ip_options_echo(struct ip_options *
dopt->is_strictroute = sopt->is_strictroute;
}
}
+ if (sopt->cipso) {
+ optlen = sptr[sopt->cipso+1];
+ dopt->cipso = dopt->optlen+sizeof(struct iphdr);
+ memcpy(dptr, sptr+sopt->cipso, optlen);
+ dptr += optlen;
+ dopt->optlen += optlen;
+ }
while (dopt->optlen & 3) {
*dptr++ = IPOPT_END;
dopt->optlen++;
@@ -435,6 +443,17 @@ int ip_options_compile(struct ip_options
if (optptr[2] == 0 && optptr[3] == 0)
opt->router_alert = optptr - iph;
break;
+ case IPOPT_CIPSO:
+ if (opt->cipso) {
+ pp_ptr = optptr;
+ goto error;
+ }
+ opt->cipso = optptr - iph;
+ if (cipso_v4_validate(&optptr)) {
+ pp_ptr = optptr;
+ goto error;
+ }
+ break;
case IPOPT_SEC:
case IPOPT_SID:
default:
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 2/7] NetLabel: core network changes
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Changes to the core network stack to support the NetLabel subsystem. This
includes changes to the IPv4 option handling to support CIPSO labels, and a new
NetLabel hook in inet_accept() to handle NetLabel attributes across accept()
calls done by in-kernel daemons.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
include/linux/ip.h | 1
include/net/cipso_ipv4.h | 266 ++++++++++++++++++++
include/net/inet_sock.h | 2
include/net/netlabel.h | 615 +++++++++++++++++++++++++++++++++++++++++++++++
net/ipv4/af_inet.c | 3
net/ipv4/ah4.c | 2
net/ipv4/ip_options.c | 19 +
7 files changed, 906 insertions(+), 2 deletions(-)
Index: linux-2.6.18-rc1/include/linux/ip.h
===================================================================
--- linux-2.6.18-rc1.orig/include/linux/ip.h
+++ linux-2.6.18-rc1/include/linux/ip.h
@@ -57,6 +57,7 @@
#define IPOPT_SEC (2 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_LSRR (3 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_TIMESTAMP (4 |IPOPT_MEASUREMENT)
+#define IPOPT_CIPSO (6 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_RR (7 |IPOPT_CONTROL)
#define IPOPT_SID (8 |IPOPT_CONTROL|IPOPT_COPY)
#define IPOPT_SSRR (9 |IPOPT_CONTROL|IPOPT_COPY)
Index: linux-2.6.18-rc1/include/net/cipso_ipv4.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/include/net/cipso_ipv4.h
@@ -0,0 +1,266 @@
+/*
+ * CIPSO - Commercial IP Security Option
+ *
+ * This is an implementation of the CIPSO 2.2 protocol as specified in
+ * draft-ietf-cipso-ipsecurity-01.txt with additional tag types as found in
+ * FIPS-188, copies of both documents can be found in the Documentation
+ * directory. While CIPSO never became a full IETF RFC standard many vendors
+ * have chosen to adopt the protocol and over the years it has become a
+ * de-facto standard for labeled networking.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _CIPSO_IPV4_H
+#define _CIPSO_IPV4_H
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <net/netlabel.h>
+
+/* known doi values */
+#define CIPSO_V4_DOI_UNKNOWN 0x00000000
+
+/* tag types */
+#define CIPSO_V4_TAG_INVALID 0
+#define CIPSO_V4_TAG_RBITMAP 1
+#define CIPSO_V4_TAG_ENUM 2
+#define CIPSO_V4_TAG_RANGE 5
+#define CIPSO_V4_TAG_PBITMAP 6
+#define CIPSO_V4_TAG_FREEFORM 7
+
+/* doi mapping types */
+#define CIPSO_V4_MAP_UNKNOWN 0
+#define CIPSO_V4_MAP_STD 1
+#define CIPSO_V4_MAP_PASS 2
+
+/* limits */
+#define CIPSO_V4_MAX_REM_LVLS 256
+#define CIPSO_V4_INV_LVL 0x80000000
+#define CIPSO_V4_MAX_LOC_LVLS (CIPSO_V4_INV_LVL - 1)
+#define CIPSO_V4_MAX_REM_CATS 65536
+#define CIPSO_V4_INV_CAT 0x80000000
+#define CIPSO_V4_MAX_LOC_CATS (CIPSO_V4_INV_CAT - 1)
+
+/*
+ * CIPSO DOI definitions
+ */
+
+/* DOI definition struct */
+#define CIPSO_V4_TAG_MAXCNT 5
+struct cipso_v4_doi {
+ u32 doi;
+ u32 type;
+ union {
+ struct cipso_v4_std_map_tbl *std;
+ } map;
+ u8 tags[CIPSO_V4_TAG_MAXCNT];
+
+ u32 valid;
+ struct list_head list;
+ struct rcu_head rcu;
+ struct list_head dom_list;
+};
+
+/* Standard CIPSO mapping table */
+/* NOTE: the highest order bit (i.e. 0x80000000) is an 'invalid' flag, if the
+ * bit is set then consider that value as unspecified, meaning the
+ * mapping for that particular level/category is invalid */
+struct cipso_v4_std_map_tbl {
+ struct {
+ u32 *cipso;
+ u32 *local;
+ u32 cipso_size;
+ u32 local_size;
+ } lvl;
+ struct {
+ u32 *cipso;
+ u32 *local;
+ u32 cipso_size;
+ u32 local_size;
+ } cat;
+};
+
+/*
+ * Sysctl Variables
+ */
+
+#ifdef CONFIG_NETLABEL
+extern int cipso_v4_cache_enabled;
+extern int cipso_v4_cache_bucketsize;
+extern int cipso_v4_rbm_optfmt;
+extern int cipso_v4_rbm_strictvalid;
+#endif
+
+/*
+ * Helper Functions
+ */
+
+#define CIPSO_V4_OPTEXIST(x) (IPCB(x)->opt.cipso != 0)
+#define CIPSO_V4_OPTPTR(x) ((x)->nh.raw + IPCB(x)->opt.cipso)
+
+/*
+ * DOI List Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+int cipso_v4_doi_add(struct cipso_v4_doi *doi_def);
+int cipso_v4_doi_remove(const u32 doi,
+ void (*callback) (struct rcu_head * head));
+struct cipso_v4_doi *cipso_v4_doi_getdef(const u32 doi);
+struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom);
+struct sk_buff *cipso_v4_doi_dump(const u32 doi, const size_t headroom);
+int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def, const char *domain);
+int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+ const char *domain);
+#else
+static inline int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_doi_remove(const u32 doi,
+ void (*callback) (struct rcu_head * head))
+{
+ return 0;
+}
+
+static inline struct cipso_v4_doi *cipso_v4_doi_getdef(const u32 doi)
+{
+ return NULL;
+}
+
+static inline struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
+{
+ return NULL;
+}
+
+static inline struct sk_buff *cipso_v4_doi_dump(const u32 doi,
+ const size_t headroom)
+{
+ return NULL;
+}
+
+static inline int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def,
+ const char *domain)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+ const char *domain)
+{
+ return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Label Mapping Cache Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+void cipso_v4_cache_invalidate(void);
+int cipso_v4_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr);
+#else
+static inline void cipso_v4_cache_invalidate(void)
+{
+ return;
+}
+
+static inline int cipso_v4_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Protocol Handling Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+void cipso_v4_error(struct sk_buff *skb, const int error, const u32 gateway);
+int cipso_v4_socket_setopt(struct socket *sock,
+ unsigned char *opt,
+ u32 opt_len);
+int cipso_v4_socket_setattr(const struct socket *sock,
+ const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr);
+int cipso_v4_socket_getopt(const struct socket *sock,
+ unsigned char **opt,
+ u32 *opt_len);
+int cipso_v4_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr);
+int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr);
+int cipso_v4_validate(unsigned char **option);
+#else
+static inline void cipso_v4_error(struct sk_buff *skb,
+ const int error,
+ const u32 gateway)
+{
+ return;
+}
+
+static inline int cipso_v4_socket_setopt(struct socket *sock,
+ unsigned char *opt,
+ u32 opt_len)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_setattr(const struct socket *sock,
+ const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_getopt(const struct socket *sock,
+ unsigned char **opt,
+ u32 *opt_len)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int cipso_v4_validate(unsigned char **option)
+{
+ return -ENOSYS;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif /* _CIPSO_IPV4_H */
Index: linux-2.6.18-rc1/include/net/inet_sock.h
===================================================================
--- linux-2.6.18-rc1.orig/include/net/inet_sock.h
+++ linux-2.6.18-rc1/include/net/inet_sock.h
@@ -51,7 +51,7 @@ struct ip_options {
ts_needtime:1,
ts_needaddr:1;
unsigned char router_alert;
- unsigned char __pad1;
+ unsigned char cipso;
unsigned char __pad2;
unsigned char __data[0];
};
Index: linux-2.6.18-rc1/include/net/netlabel.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/include/net/netlabel.h
@@ -0,0 +1,615 @@
+/*
+ * NetLabel System
+ *
+ * The NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_H
+#define _NETLABEL_H
+
+#include <linux/types.h>
+#include <linux/skbuff.h>
+#include <net/netlink.h>
+
+/*
+ * NetLabel - A management interface for maintaining network packet label
+ * mapping tables for explicit packet labling protocols.
+ *
+ * Network protocols such as CIPSO and RIPSO require a label translation layer
+ * to convert the label on the packet into something meaningful on the host
+ * machine. In the current Linux implementation these mapping tables live
+ * inside the kernel; NetLabel provides a mechanism for user space applications
+ * to manage these mapping tables.
+ *
+ * NetLabel makes use of the Generic NETLINK mechanism as a transport layer to
+ * send messages between kernel and user space. The general format of a
+ * NetLabel message is shown below:
+ *
+ * +-----------------+-------------------+--------- --- -- -
+ * | struct nlmsghdr | struct genlmsghdr | payload
+ * +-----------------+-------------------+--------- --- -- -
+ *
+ * The 'nlmsghdr' and 'genlmsghdr' structs should be dealt with like normal.
+ * The payload is dependent on the subsystem specified in the
+ * 'nlmsghdr->nlmsg_type' and should be defined below, supporting functions
+ * should be defined in the corresponding net/netlabel/netlabel_<subsys>.h|c
+ * file. All of the fields in the NetLabel payload should be aligned using
+ * the alignment functions provided.
+ *
+ */
+
+/*
+ * NetLabel NETLINK protocol
+ */
+
+#define NETLBL_PROTO_VERSION 1
+
+/* NetLabel NETLINK types/families */
+#define NETLBL_NLTYPE_NONE 0
+#define NETLBL_NLTYPE_MGMT 1
+#define NETLBL_NLTYPE_MGMT_NAME "NLBL_MGMT"
+#define NETLBL_NLTYPE_RIPSO 2
+#define NETLBL_NLTYPE_RIPSO_NAME "NLBL_RIPSO"
+#define NETLBL_NLTYPE_CIPSOV4 3
+#define NETLBL_NLTYPE_CIPSOV4_NAME "NLBL_CIPSOv4"
+#define NETLBL_NLTYPE_CIPSOV6 4
+#define NETLBL_NLTYPE_CIPSOV6_NAME "NLBL_CIPSOv6"
+#define NETLBL_NLTYPE_UNLABELED 5
+#define NETLBL_NLTYPE_UNLABELED_NAME "NLBL_UNLBL"
+
+/* NetLabel return codes */
+#define NETLBL_E_OK 0
+
+/*
+ * Helper functions
+ */
+
+#define NETLBL_LEN_U8 netlbl_align(1)
+#define NETLBL_LEN_U16 netlbl_align(2)
+#define NETLBL_LEN_U32 netlbl_align(4)
+/**
+ * netlbl_align - Align a NetLabel data chunk
+ * @len: the data chunk length
+ *
+ * Description:
+ * Return the aligned data chunk length.
+ *
+ */
+static inline size_t netlbl_align(const size_t length)
+{
+ return NLMSG_ALIGN(length);
+}
+
+/**
+ * netlbl_put_u8 - Write a u8 value into a buffer
+ * @buffer: the buffer
+ * @val: the value
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer.
+ *
+ */
+static inline void netlbl_put_u8(unsigned char *buffer, const u8 val)
+{
+ *(u8 *)buffer = val;
+}
+
+/**
+ * netlbl_put_u16 - Write a u16 value into a buffer
+ * @buffer: the buffer
+ * @val: the value
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer.
+ *
+ */
+static inline void netlbl_put_u16(unsigned char *buffer, const u16 val)
+{
+ *(u16 *)buffer = val;
+}
+
+/**
+ * netlbl_put_u32 - Write a u32 value into a buffer
+ * @buffer: the buffer
+ * @val: the value
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer.
+ *
+ */
+static inline void netlbl_put_u32(unsigned char *buffer, const u32 val)
+{
+ *(u32 *)buffer = val;
+}
+
+/**
+ * netlbl_put_hdr - Write a NETLINK header into a buffer
+ * @buffer: the buffer
+ * @msg_type: the NETLINK message type
+ * @msg_len: the NETLINK message length
+ * @msg_flags: the NETLINK message flags
+ * @msg_pid: the NETLINK message PID
+ * @msg_seq: the NETLINK message sequence number
+ *
+ * Description:
+ * Use the given values to write a NETLINK header into the given buffer.
+ *
+ */
+static inline void netlbl_put_hdr(unsigned char *buffer,
+ const u32 msg_type,
+ const u16 msg_len,
+ const u16 msg_flags,
+ const u32 msg_pid,
+ const u32 msg_seq)
+{
+ struct nlmsghdr *hdr = (struct nlmsghdr *)buffer;
+ hdr->nlmsg_len = msg_len;
+ hdr->nlmsg_type = msg_type;
+ hdr->nlmsg_flags = msg_flags;
+ hdr->nlmsg_seq = msg_seq;
+ hdr->nlmsg_pid = msg_pid;
+}
+
+/**
+ * netlbl_putinc_u8 - Write a u8 value into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_u8(unsigned char **buffer,
+ const u8 val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u8));
+ netlbl_put_u8(*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_putinc_u16 - Write a u16 value into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_u16(unsigned char **buffer,
+ const u16 val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u16));
+ netlbl_put_u16(*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_putinc_u32 - Write a u32 value into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the value specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_u32(unsigned char **buffer,
+ const u32 val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u32));
+ netlbl_put_u32(*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_putinc_str - Write a string into a buffer and increment the buffer
+ * @buffer: the buffer
+ * @val: the value
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Write the string specified in @val into the buffer specified by @buffer
+ * and advance the buffer pointer past the newly written value. If @rem_len
+ * is not NULL then decrement it by the field length.
+ *
+ */
+static inline void netlbl_putinc_str(unsigned char **buffer,
+ const char *val,
+ ssize_t *rem_len)
+{
+ size_t len = netlbl_align(strlen(val) + 1);
+ strcpy((char *)*buffer, val);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+}
+
+/**
+ * netlbl_put_hdr - Write a NETLINK header into a buffer and increment the ptr
+ * @buffer: the buffer
+ * @msg_type: the NETLINK message type
+ * @msg_len: the NETLINK message length
+ * @msg_flags: the NETLINK message flags
+ * @msg_pid: the NETLINK message PID
+ * @msg_seq: the NETLINK message sequence number
+ *
+ * Description:
+ * Use the given values to write a NETLINK header into the given buffer and
+ * then increment the buffer pointer past the header.
+ *
+ */
+static inline void netlbl_putinc_hdr(unsigned char **buffer,
+ const u32 msg_type,
+ const u16 msg_len,
+ const u16 msg_flags,
+ const u32 msg_pid,
+ const u32 msg_seq)
+{
+ netlbl_put_hdr(*buffer,
+ msg_type,
+ msg_len,
+ msg_flags,
+ msg_pid,
+ msg_seq);
+ *buffer += NLMSG_HDRLEN;
+}
+
+/**
+ * netlbl_get_u8 - Read a u8 value from a buffer
+ * @buffer: the buffer
+ *
+ * Description:
+ * Return a u8 value pointed to by @buffer.
+ *
+ */
+static inline u8 netlbl_get_u8(const unsigned char *buffer)
+{
+ return *(u8 *)buffer;
+}
+
+/**
+ * netlbl_get_u16 - Read a u16 value from a buffer
+ * @buffer: the buffer
+ *
+ * Description:
+ * Return a u16 value pointed to by @buffer.
+ *
+ */
+static inline u16 netlbl_get_u16(const unsigned char *buffer)
+{
+ return *(u16 *)buffer;
+}
+
+/**
+ * netlbl_get_u32 - Read a u32 value from a buffer
+ * @buffer: the buffer
+ *
+ * Description:
+ * Return a u32 value pointed to by @buffer.
+ *
+ */
+static inline u32 netlbl_get_u32(const unsigned char *buffer)
+{
+ return *(u32 *)buffer;
+}
+
+/**
+ * netlbl_getinc_u8 - Read a u8 value from a buffer and increment the buffer
+ * @buffer: the buffer
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Return a u8 value pointed to by @buffer and increment the buffer pointer
+ * past the value. If @rem_len is not NULL, decrement it by the field size.
+ *
+ */
+static inline u8 netlbl_getinc_u8(unsigned char **buffer, ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u8));
+ u8 val = netlbl_get_u8(*buffer);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+ return val;
+}
+
+/**
+ * netlbl_getinc_u16 - Read a u16 value from a buffer and increment the buffer
+ * @buffer: the buffer
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Return a u16 value pointed to by @buffer and increment the buffer pointer
+ * past the value. If @rem_len is not NULL, decrement it by the field size.
+ *
+ */
+static inline u16 netlbl_getinc_u16(unsigned char **buffer, ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u16));
+ u16 val = netlbl_get_u16(*buffer);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+ return val;
+}
+
+/**
+ * netlbl_getinc_u32 - Read a u32 value from a buffer and increment the buffer
+ * @buffer: the buffer
+ * @rem_len: remaining length
+ *
+ * Description:
+ * Return a u32 value pointed to by @buffer and increment the buffer pointer
+ * past the value. If @rem_len is not NULL, decrement it by the field size.
+ *
+ */
+static inline u32 netlbl_getinc_u32(unsigned char **buffer, ssize_t *rem_len)
+{
+ size_t len = netlbl_align(sizeof(u32));
+ u32 val = netlbl_get_u32(*buffer);
+ *buffer += len;
+ if (rem_len != NULL)
+ *rem_len -= len;
+ return val;
+}
+
+/**
+ * netlbl_netlink_alloc_skb - Allocate a NETLINK message buffer
+ * @head: the amount of headroom in bytes
+ * @body: the desired size (minus headroom) in bytes
+ * @gfp_flags: the alloc flags to pass to alloc_skb()
+ *
+ * Description:
+ * Allocate a NETLINK message buffer based on the sizes given in @head and
+ * @body. If @head is greater than zero skb_reserve() is called to reserve
+ * @head bytes at the start of the buffer. Returns a valid sk_buff pointer on
+ * success, NULL on failure.
+ *
+ */
+static inline struct sk_buff *netlbl_netlink_alloc_skb(const size_t head,
+ const size_t body,
+ int gfp_flags)
+{
+ struct sk_buff *skb;
+
+ skb = alloc_skb(NLMSG_ALIGN(head + body), gfp_flags);
+ if (skb == NULL)
+ return NULL;
+ if (head > 0) {
+ skb_reserve(skb, head);
+ if (skb_tailroom(skb) < body) {
+ kfree_skb(skb);
+ return NULL;
+ }
+ }
+
+ return skb;
+}
+
+/*
+ * NetLabel - Kernel API for accessing the network packet label mappings.
+ *
+ * The following functions are provided for use by other kernel modules,
+ * specifically kernel LSM modules, to provide a consistent, transparent API
+ * for dealing with explicit packet labeling protocols such as CIPSO and
+ * RIPSO. The functions defined here are implemented in the
+ * net/netlabel/netlabel_kapi.c file.
+ *
+ */
+
+/* Domain mapping definition struct */
+struct netlbl_dom_map;
+
+/* Domain mapping operations */
+int netlbl_domhsh_remove(const char *domain);
+
+/* LSM security attributes */
+struct netlbl_lsm_cache {
+ void (*free) (const void *data);
+ void *data;
+};
+struct netlbl_lsm_secattr {
+ char *domain;
+
+ u32 mls_lvl;
+ u32 mls_lvl_vld;
+ unsigned char *mls_cat;
+ size_t mls_cat_len;
+
+ struct netlbl_lsm_cache cache;
+};
+
+/*
+ * LSM security attribute operations
+ */
+
+
+/**
+ * netlbl_secattr_init - Initialize a netlbl_lsm_secattr struct
+ * @secattr: the struct to initialize
+ *
+ * Description:
+ * Initialize an already allocated netlbl_lsm_secattr struct. Returns zero on
+ * success, negative values on error.
+ *
+ */
+static inline int netlbl_secattr_init(struct netlbl_lsm_secattr *secattr)
+{
+ memset(secattr, 0, sizeof(*secattr));
+ return 0;
+}
+
+/**
+ * netlbl_secattr_destroy - Clears a netlbl_lsm_secattr struct
+ * @secattr: the struct to clear
+ * @clear_cache: cache clear flag
+ *
+ * Description:
+ * Destroys the @secattr struct, including freeing all of the internal buffers.
+ * If @clear_cache is true then free the cache fields, otherwise leave them
+ * intact. The struct must be reset with a call to netlbl_secattr_init()
+ * before reuse.
+ *
+ */
+static inline void netlbl_secattr_destroy(struct netlbl_lsm_secattr *secattr,
+ const u32 clear_cache)
+{
+ if (clear_cache && secattr->cache.data != NULL && secattr->cache.free)
+ secattr->cache.free(secattr->cache.data);
+ kfree(secattr->domain);
+ kfree(secattr->mls_cat);
+}
+
+/**
+ * netlbl_secattr_alloc - Allocate and initialize a netlbl_lsm_secattr struct
+ * @flags: the memory allocation flags
+ *
+ * Description:
+ * Allocate and initialize a netlbl_lsm_secattr struct. Returns a valid
+ * pointer on success, or NULL on failure.
+ *
+ */
+static inline struct netlbl_lsm_secattr *netlbl_secattr_alloc(const int flags)
+{
+ return kzalloc(sizeof(struct netlbl_lsm_secattr), flags);
+}
+
+/**
+ * netlbl_secattr_free - Frees a netlbl_lsm_secattr struct
+ * @secattr: the struct to free
+ * @clear_cache: cache clear flag
+ *
+ * Description:
+ * Frees @secattr including all of the internal buffers. If @clear_cache is
+ * true then free the cache fields, otherwise leave them intact.
+ *
+ */
+static inline void netlbl_secattr_free(struct netlbl_lsm_secattr *secattr,
+ const u32 clear_cache)
+{
+ netlbl_secattr_destroy(secattr, clear_cache);
+ kfree(secattr);
+}
+
+/*
+ * LSM protocol operations
+ */
+
+#ifdef CONFIG_NETLABEL
+int netlbl_socket_setattr(const struct socket *sock,
+ const struct netlbl_lsm_secattr *secattr);
+int netlbl_socket_peekattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr);
+int netlbl_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr);
+int netlbl_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr);
+void netlbl_skbuff_err(struct sk_buff *skb, int error);
+#else
+static inline int netlbl_socket_setattr(const struct socket *sock,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int netlbl_socket_peekattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int netlbl_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline int netlbl_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ return -ENOSYS;
+}
+
+static inline void netlbl_skbuff_err(struct sk_buff *skb, int error)
+{
+ return;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * LSM label mapping cache operations
+ */
+
+#ifdef CONFIG_NETLABEL
+void netlbl_cache_invalidate(void);
+int netlbl_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr);
+#else
+static inline void netlbl_cache_invalidate(void)
+{
+ return;
+}
+
+static inline int netlbl_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Network stack operations
+ */
+
+#ifdef CONFIG_NETLABEL
+void netlbl_socket_inet_accept(struct socket *sock, struct socket *newsock);
+#else
+static inline void netlbl_socket_inet_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ return;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif /* _NETLABEL_H */
Index: linux-2.6.18-rc1/net/ipv4/af_inet.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/af_inet.c
+++ linux-2.6.18-rc1/net/ipv4/af_inet.c
@@ -115,6 +115,7 @@
#ifdef CONFIG_IP_MROUTE
#include <linux/mroute.h>
#endif
+#include <net/netlabel.h>
DEFINE_SNMP_STAT(struct linux_mib, net_statistics) __read_mostly;
@@ -617,6 +618,8 @@ int inet_accept(struct socket *sock, str
sock_graft(sk2, newsock);
+ netlbl_socket_inet_accept(sock, newsock);
+
newsock->state = SS_CONNECTED;
err = 0;
release_sock(sk2);
Index: linux-2.6.18-rc1/net/ipv4/ah4.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/ah4.c
+++ linux-2.6.18-rc1/net/ipv4/ah4.c
@@ -34,7 +34,7 @@ static int ip_clear_mutable_options(stru
switch (*optptr) {
case IPOPT_SEC:
case 0x85: /* Some "Extended Security" crap. */
- case 0x86: /* Another "Commercial Security" crap. */
+ case IPOPT_CIPSO:
case IPOPT_RA:
case 0x80|21: /* RFC1770 */
break;
Index: linux-2.6.18-rc1/net/ipv4/ip_options.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/ip_options.c
+++ linux-2.6.18-rc1/net/ipv4/ip_options.c
@@ -24,6 +24,7 @@
#include <net/ip.h>
#include <net/icmp.h>
#include <net/route.h>
+#include <net/cipso_ipv4.h>
/*
* Write options to IP header, record destination address to
@@ -194,6 +195,13 @@ int ip_options_echo(struct ip_options *
dopt->is_strictroute = sopt->is_strictroute;
}
}
+ if (sopt->cipso) {
+ optlen = sptr[sopt->cipso+1];
+ dopt->cipso = dopt->optlen+sizeof(struct iphdr);
+ memcpy(dptr, sptr+sopt->cipso, optlen);
+ dptr += optlen;
+ dopt->optlen += optlen;
+ }
while (dopt->optlen & 3) {
*dptr++ = IPOPT_END;
dopt->optlen++;
@@ -435,6 +443,17 @@ int ip_options_compile(struct ip_options
if (optptr[2] == 0 && optptr[3] == 0)
opt->router_alert = optptr - iph;
break;
+ case IPOPT_CIPSO:
+ if (opt->cipso) {
+ pp_ptr = optptr;
+ goto error;
+ }
+ opt->cipso = optptr - iph;
+ if (cipso_v4_validate(&optptr)) {
+ pp_ptr = optptr;
+ goto error;
+ }
+ break;
case IPOPT_SEC:
case IPOPT_SID:
default:
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-net_cipsov4-2.6.18 --]
[-- Type: text/plain, Size: 52030 bytes --]
Add support for the Commercial IP Security Option (CIPSO) to the IPv4 network
stack. CIPSO has become a de-facto standard for trusted/labeled networking
amongst existing Trusted Operating Systems such as Trusted Solaris, HP-UX CMW,
etc. This implementation is designed to be used with the NetLabel subsystem
to provide explicit packet labeling to LSM developers.
The CIPSO/IPv4 packet labeling works by the LSM calling a NetLabel API function
which attaches a CIPSO label (IPv4 option) to a given socket; this in turn
attaches the CIPSO label to every packet leaving the socket without any extra
processing on the outbound side. On the inbound side the individual packet's
sk_buff is examined through a call to a NetLabel API function to determine if a
CIPSO/IPv4 label is present and if so the security attributes of the CIPSO
label are returned to the caller of the NetLabel API function.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
include/linux/sysctl.h | 4
net/ipv4/Makefile | 1
net/ipv4/cipso_ipv4.c | 1708 +++++++++++++++++++++++++++++++++++++++++++++
net/ipv4/sysctl_net_ipv4.c | 35
4 files changed, 1748 insertions(+)
Index: linux-2.6.18-rc1/include/linux/sysctl.h
===================================================================
--- linux-2.6.18-rc1.orig/include/linux/sysctl.h
+++ linux-2.6.18-rc1/include/linux/sysctl.h
@@ -411,6 +411,10 @@ enum
NET_IPV4_TCP_WORKAROUND_SIGNED_WINDOWS=115,
NET_TCP_DMA_COPYBREAK=116,
NET_TCP_SLOW_START_AFTER_IDLE=117,
+ NET_CIPSOV4_CACHE_ENABLE=118,
+ NET_CIPSOV4_CACHE_BUCKET_SIZE=119,
+ NET_CIPSOV4_RBM_OPTFMT=120,
+ NET_CIPSOV4_RBM_STRICTVALID=121,
};
enum {
Index: linux-2.6.18-rc1/net/ipv4/Makefile
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/Makefile
+++ linux-2.6.18-rc1/net/ipv4/Makefile
@@ -48,6 +48,7 @@ obj-$(CONFIG_TCP_CONG_VENO) += tcp_veno.
obj-$(CONFIG_TCP_CONG_SCALABLE) += tcp_scalable.o
obj-$(CONFIG_TCP_CONG_LP) += tcp_lp.o
obj-$(CONFIG_TCP_CONG_COMPOUND) += tcp_compound.o
+obj-$(CONFIG_NETLABEL) += cipso_ipv4.o
obj-$(CONFIG_XFRM) += xfrm4_policy.o xfrm4_state.o xfrm4_input.o \
xfrm4_output.o
Index: linux-2.6.18-rc1/net/ipv4/cipso_ipv4.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/ipv4/cipso_ipv4.c
@@ -0,0 +1,1708 @@
+/*
+ * CIPSO - Commercial IP Security Option
+ *
+ * This is an implementation of the CIPSO 2.2 protocol as specified in
+ * draft-ietf-cipso-ipsecurity-01.txt with additional tag types as found in
+ * FIPS-188, copies of both documents can be found in the Documentation
+ * directory. While CIPSO never became a full IETF RFC standard many vendors
+ * have chosen to adopt the protocol and over the years it has become a
+ * de-facto standard for labeled networking.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <linux/module.h>
+#include <linux/jhash.h>
+#include <net/ip.h>
+#include <net/icmp.h>
+#include <net/tcp.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+struct cipso_v4_domhsh_entry {
+ char *domain;
+ u32 valid;
+ struct list_head list;
+ struct rcu_head rcu;
+};
+
+/* List of available DOI definitions */
+/* XXX - Updates should be minimal so having a single lock for the
+ cipso_v4_doi_list and the cipso_v4_doi_list->dom_list should be
+ okay. */
+/* XXX - This currently assumes a minimal number of different DOIs in use,
+ if in practice there are a lot of different DOIs this list should
+ probably be turned into a hash table or something similar so we
+ can do quick lookups. */
+DEFINE_SPINLOCK(cipso_v4_doi_list_lock);
+static struct list_head cipso_v4_doi_list = LIST_HEAD_INIT(cipso_v4_doi_list);
+
+/* Label mapping cache */
+int cipso_v4_cache_enabled = 1;
+int cipso_v4_cache_bucketsize = 10;
+#define CIPSO_V4_CACHE_BUCKETBITS 7
+#define CIPSO_V4_CACHE_BUCKETS (1 << CIPSO_V4_CACHE_BUCKETBITS)
+#define CIPSO_V4_CACHE_REORDERLIMIT 10
+struct cipso_v4_map_cache_bkt {
+ spinlock_t lock;
+ u32 size;
+ struct list_head list;
+};
+struct cipso_v4_map_cache_entry {
+ u32 hash;
+ unsigned char *key;
+ u32 key_len;
+
+ struct netlbl_lsm_cache lsm_data;
+
+ u32 activity;
+ struct list_head list;
+};
+static u32 cipso_v4_cache_size = 0;
+static struct cipso_v4_map_cache_bkt *cipso_v4_cache = NULL;
+#define CIPSO_V4_CACHE_ENABLED (cipso_v4_cache_enabled && \
+ cipso_v4_cache_bucketsize > 0 && \
+ cipso_v4_cache_size > 0)
+
+/* Restricted bitmap (tag #1) flags */
+int cipso_v4_rbm_optfmt = 0;
+int cipso_v4_rbm_strictvalid = 1;
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
+ * @bitmap: the bitmap
+ * @bitmap_len: length in bits
+ * @offset: starting offset
+ * @state: if non-zero, look for a set (1) bit else look for a cleared (0) bit
+ *
+ * Description:
+ * Starting at @offset, walk the bitmap from left to right until either the
+ * desired bit is found or we reach the end. Return the bit offset, -1 if
+ * not found, or -2 if error.
+ */
+static int cipso_v4_bitmap_walk(const unsigned char *bitmap,
+ const u32 bitmap_len,
+ const u32 offset,
+ const u8 state)
+{
+ u32 bit_spot;
+ u32 byte_offset;
+ unsigned char bitmask;
+ unsigned char byte;
+
+ /* gcc always rounds to zero when doing integer division */
+ byte_offset = offset / 8;
+ byte = bitmap[byte_offset];
+ bit_spot = offset;
+ bitmask = 0x80 >> (offset % 8);
+
+ /* XXX - probably should use include/asm/bitops.h if we can */
+ while (bit_spot < bitmap_len) {
+ if ((state && (byte & bitmask) == bitmask) ||
+ (state == 0 && (byte & bitmask) == 0))
+ return bit_spot;
+
+ bit_spot++;
+ bitmask >>= 1;
+ if (bitmask == 0) {
+ byte = bitmap[++byte_offset];
+ bitmask = 0x80;
+ }
+ }
+
+ return -1;
+}
+
+/**
+ * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
+ * @bitmap: the bitmap
+ * @bit: the bit
+ * @state: if non-zero, set the bit (1) else clear the bit (0)
+ *
+ * Description:
+ * Set a single bit in the bitmask. Returns zero on success, negative values
+ * on error.
+ */
+static void cipso_v4_bitmap_setbit(unsigned char *bitmap,
+ const u32 bit,
+ const u8 state)
+{
+ u32 byte_spot;
+ u8 bitmask;
+
+ /* gcc always rounds to zero when doing integer division */
+ byte_spot = bit / 8;
+ bitmask = 0x80 >> (bit % 8);
+ if (state)
+ bitmap[byte_spot] |= bitmask;
+ else
+ bitmap[byte_spot] &= ~bitmask;
+}
+
+/**
+ * cipso_v4_doi_domhsh_free - Frees a domain list entry
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to a domain list entry can be released
+ * safely.
+ *
+ */
+static void cipso_v4_doi_domhsh_free(struct rcu_head *entry)
+{
+ struct cipso_v4_domhsh_entry *ptr;
+
+ ptr = container_of(entry, struct cipso_v4_domhsh_entry, rcu);
+ kfree(ptr->domain);
+ kfree(ptr);
+}
+
+/**
+ * cipso_v4_cache_entry_free - Frees a cache entry
+ * @entry: the entry to free
+ *
+ * Description:
+ * This function frees the memory associated with a cache entry.
+ *
+ */
+static void cipso_v4_cache_entry_free(struct cipso_v4_map_cache_entry *entry)
+{
+ if (entry->lsm_data.free)
+ entry->lsm_data.free(entry->lsm_data.data);
+ kfree(entry->key);
+ kfree(entry);
+}
+
+/**
+ * cipso_v4_map_cache_hash - Hashing function for the CIPSO cache
+ * @key: the hash key
+ * @key_len: the length of the key in bytes
+ *
+ * Description:
+ * The CIPSO tag hashing function. Returns a 32-bit hash value.
+ *
+ */
+static u32 cipso_v4_map_cache_hash(const unsigned char *key, const u32 key_len)
+{
+ return jhash(key, key_len, 0);
+}
+
+/*
+ * Label Mapping Cache Functions
+ */
+
+/**
+ * cipso_v4_cache_init - Initialize the CIPSO cache
+ * @bkt_size: the number of cache buckets
+ *
+ * Description:
+ * Initializes the CIPSO label mapping cache, this function should be called
+ * before any of the other functions defined in this file. Returns zero on
+ * success, negative values on error.
+ *
+ */
+static int cipso_v4_cache_init(const u32 bkt_size)
+{
+ struct cipso_v4_map_cache_bkt *cache;
+ u32 iter;
+
+ if (bkt_size == 0)
+ return -EINVAL;
+
+ cache = kcalloc(bkt_size,
+ sizeof(struct cipso_v4_map_cache_bkt), GFP_KERNEL);
+ if (cache == NULL)
+ return -ENOMEM;
+
+ for (iter = 0; iter < bkt_size; iter++) {
+ cache[iter].lock = SPIN_LOCK_UNLOCKED;
+ cache[iter].size = 0;
+ INIT_LIST_HEAD(&cache[iter].list);
+ }
+ cipso_v4_cache = cache;
+ cipso_v4_cache_size = bkt_size;
+
+ return 0;
+}
+
+/**
+ * cipso_v4_cache_destroy - Destroy the CIPSO cache
+ *
+ * Description:
+ * Clears the CIPSO cache and frees all the memory. This function does not
+ * hold any locks and should only be called when the module is being unloaded.
+ *
+ */
+static int cipso_v4_cache_destroy(void)
+{
+ struct cipso_v4_map_cache_bkt *cache;
+ struct cipso_v4_map_cache_entry *entry;
+ u32 cache_size;
+ u32 iter;
+
+ cache = cipso_v4_cache;
+ cache_size = cipso_v4_cache_size;
+ cipso_v4_cache_size = 0;
+ cipso_v4_cache = NULL;
+
+ for (iter = 0; iter < cache_size; iter++)
+ list_for_each_entry(entry, &cache[iter].list, list) {
+ list_del(&entry->list);
+ cipso_v4_cache_entry_free(entry);
+ }
+
+ return 0;
+}
+
+/**
+ * cipso_v4_cache_invalidate - Invalidates the current CIPSO cache
+ *
+ * Description:
+ * Invalidates and frees any entries in the CIPSO cache. Returns zero on
+ * success and negative values on failure.
+ *
+ */
+void cipso_v4_cache_invalidate(void)
+{
+ struct cipso_v4_map_cache_entry *entry, *tmp_entry;
+ u32 iter;
+
+ if (cipso_v4_cache_size == 0)
+ return;
+
+ for (iter = 0; iter < cipso_v4_cache_size; iter++) {
+ spin_lock(&cipso_v4_cache[iter].lock);
+ list_for_each_entry_safe(entry,
+ tmp_entry,
+ &cipso_v4_cache[iter].list, list) {
+ list_del(&entry->list);
+ cipso_v4_cache_entry_free(entry);
+ }
+ cipso_v4_cache[iter].size = 0;
+ spin_unlock(&cipso_v4_cache[iter].lock);
+ }
+
+ return;
+}
+
+/**
+ * cipso_v4_cache_check - Check the CIPSO cache for a label mapping
+ * @key: the buffer to check
+ * @key_len: buffer length in bytes
+ * @secattr: the security attribute struct to use
+ *
+ * Description:
+ * This function checks the cache to see if a label mapping already exists for
+ * the given key. If there is a match then the cache is adjusted and the
+ * @secattr struct is populated with the correct LSM security attributes. The
+ * cache is adjusted in the following manner if the entry is not already the
+ * first in the cache bucket:
+ *
+ * 1. The cache entry's activity counter is incremented
+ * 2. The previous (higher ranking) entry's activity counter is decremented
+ * 3. If the difference between the two activity counters is geater than
+ * CIPSO_V4_CACHE_REORDERLIMIT the two entries are swapped
+ *
+ * Returns zero on success, -ENOENT for a cache miss, and other negative values
+ * on error.
+ *
+ */
+static int cipso_v4_cache_check(const unsigned char *key,
+ const u32 key_len,
+ struct netlbl_lsm_secattr *secattr)
+{
+ u32 bkt;
+ struct cipso_v4_map_cache_entry *entry;
+ struct cipso_v4_map_cache_entry *prev_entry = NULL;
+ u32 hash;
+
+ if (!CIPSO_V4_CACHE_ENABLED)
+ return -ENOENT;
+
+ hash = cipso_v4_map_cache_hash(key, key_len);
+ bkt = hash & (CIPSO_V4_CACHE_BUCKETBITS - 1);
+ spin_lock(&cipso_v4_cache[bkt].lock);
+ list_for_each_entry(entry, &cipso_v4_cache[bkt].list, list) {
+ if (entry->hash == hash &&
+ entry->key_len == key_len &&
+ memcmp(entry->key, key, key_len) == 0) {
+ entry->activity += 1;
+
+ if (prev_entry != NULL) {
+ if (prev_entry->activity > 0)
+ prev_entry->activity -= 1;
+
+ if (entry->activity > prev_entry->activity &&
+ entry->activity - prev_entry->activity >
+ CIPSO_V4_CACHE_REORDERLIMIT) {
+ __list_del(entry->list.prev,
+ entry->list.next);
+ __list_add(&entry->list,
+ prev_entry->list.prev,
+ &prev_entry->list);
+ }
+ }
+
+ secattr->cache.free = entry->lsm_data.free;
+ secattr->cache.data = entry->lsm_data.data;
+ spin_unlock(&cipso_v4_cache[bkt].lock);
+ return 0;
+ }
+ prev_entry = entry;
+ }
+ spin_unlock(&cipso_v4_cache[bkt].lock);
+
+ return -ENOENT;
+}
+
+/**
+ * cipso_v4_cache_add - Add an entry to the CIPSO cache
+ * @skb: the packet
+ * @secattr: the packet's security attributes
+ *
+ * Description:
+ * Add a new entry into the CIPSO label mapping cache. Add the new entry to
+ * head of the cache bucket's list, if the cache bucket is out of room remove
+ * the last entry in the list first. It is important to note that there is
+ * currently no checking for duplicate keys. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int cipso_v4_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -EPERM;
+ u32 bkt;
+ struct cipso_v4_map_cache_entry *entry = NULL;
+ struct cipso_v4_map_cache_entry *old_entry = NULL;
+ unsigned char *cipso_ptr;
+ u32 cipso_ptr_len;
+
+ if (!CIPSO_V4_CACHE_ENABLED)
+ return 0;
+
+ cipso_ptr = CIPSO_V4_OPTPTR(skb);
+ cipso_ptr_len = cipso_ptr[1];
+
+ entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
+ if (entry == NULL)
+ return -ENOMEM;
+ entry->key = kmalloc(cipso_ptr_len, GFP_ATOMIC);
+ if (entry->key == NULL) {
+ ret_val = -ENOMEM;
+ goto cache_add_failure;
+ }
+ memcpy(entry->key, cipso_ptr, cipso_ptr_len);
+ entry->key_len = cipso_ptr_len;
+ entry->hash = cipso_v4_map_cache_hash(cipso_ptr, cipso_ptr_len);
+ entry->lsm_data.free = secattr->cache.free;
+ entry->lsm_data.data = secattr->cache.data;
+
+ bkt = entry->hash & (CIPSO_V4_CACHE_BUCKETBITS - 1);
+ spin_lock(&cipso_v4_cache[bkt].lock);
+ if (cipso_v4_cache[bkt].size < cipso_v4_cache_bucketsize) {
+ list_add(&entry->list, &cipso_v4_cache[bkt].list);
+ cipso_v4_cache[bkt].size += 1;
+ } else {
+ old_entry = list_entry(cipso_v4_cache[bkt].list.prev,
+ struct cipso_v4_map_cache_entry, list);
+ list_del(&old_entry->list);
+ list_add(&entry->list, &cipso_v4_cache[bkt].list);
+ cipso_v4_cache_entry_free(old_entry);
+ }
+ spin_unlock(&cipso_v4_cache[bkt].lock);
+
+ return 0;
+
+cache_add_failure:
+ if (entry)
+ cipso_v4_cache_entry_free(entry);
+ return ret_val;
+}
+
+/*
+ * DOI List Functions
+ */
+
+/**
+ * cipso_v4_doi_search - Searches for a DOI definition
+ * @doi: the DOI to search for
+ *
+ * Description:
+ * Search the DOI definition list for a DOI definition with a DOI value that
+ * matches @doi. The caller is responsibile for calling rcu_read_[un]lock().
+ * Returns a pointer to the DOI definition on success and NULL on failure.
+ */
+static struct cipso_v4_doi *cipso_v4_doi_search(const u32 doi)
+{
+ struct cipso_v4_doi *iter;
+
+ list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+ if (iter->doi == doi && iter->valid)
+ return iter;
+ return NULL;
+}
+
+/**
+ * cipso_v4_doi_add - Add a new DOI to the CIPSO protocol engine
+ * @doi_def: the DOI structure
+ *
+ * Description:
+ * The caller defines a new DOI for use by the CIPSO engine and calls this
+ * function to add it to the list of acceptable domains. The caller must
+ * ensure that the mapping table specified in @doi_def->map meets all of the
+ * requirements of the mapping type (see cipso_ipv4.h for details). Returns
+ * zero on success and non-zero on failure.
+ *
+ */
+int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
+{
+ if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
+ return -EINVAL;
+
+ doi_def->valid = 1;
+ INIT_RCU_HEAD(&doi_def->rcu);
+ INIT_LIST_HEAD(&doi_def->dom_list);
+
+ rcu_read_lock();
+ if (cipso_v4_doi_search(doi_def->doi) != NULL) {
+ rcu_read_unlock();
+ return -EEXIST;
+ }
+ spin_lock(&cipso_v4_doi_list_lock);
+ list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * cipso_v4_doi_remove - Remove an existing DOI from the CIPSO protocol engine
+ * @doi: the DOI value
+ * @callback: the DOI cleanup/free callback
+ *
+ * Description:
+ * Removes a DOI definition from the CIPSO engine, @callback is called to
+ * free any memory. The NetLabel routines will be called to release their own
+ * LSM domain mappings as well as our own domain list. Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_doi_remove(const u32 doi,
+ void (*callback) (struct rcu_head * head))
+{
+ struct cipso_v4_doi *doi_def;
+ struct cipso_v4_domhsh_entry *dom_iter;
+
+ rcu_read_lock();
+ doi_def = cipso_v4_doi_search(doi);
+ if (doi_def != NULL) {
+ spin_lock(&cipso_v4_doi_list_lock);
+ doi_def->valid = 0;
+ list_del_rcu(&doi_def->list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ list_for_each_entry_rcu(dom_iter, &doi_def->dom_list, list)
+ if (dom_iter->valid)
+ netlbl_domhsh_remove(dom_iter->domain);
+ cipso_v4_cache_invalidate();
+ rcu_read_unlock();
+
+ call_rcu(&doi_def->rcu, callback);
+ return 0;
+ }
+ rcu_read_unlock();
+
+ return -ENOENT;
+}
+
+/**
+ * cipso_v4_doi_getdef - Returns a pointer to a valid DOI definition
+ * @doi: the DOI value
+ *
+ * Description:
+ * Searches for a valid DOI definition and if one is found it is returned to
+ * the caller. Otherwise NULL is returned. The caller must ensure that
+ * rcu_read_lock() is held while accessing the returned definition.
+ *
+ */
+struct cipso_v4_doi *cipso_v4_doi_getdef(const u32 doi)
+{
+ return cipso_v4_doi_search(doi);
+}
+
+/**
+ * cipso_v4_doi_dump_all - Dump all the CIPSO DOI definitions into a sk_buff
+ * @headroom: the amount of headroom to allocate for the sk_buff
+ *
+ * Description:
+ * Dump a list of all the configured DOI values into a sk_buff. The returned
+ * sk_buff has room at the front of the sk_buff for @headroom bytes. See
+ * net/netlabel/netlabel_cipso_v4.h for the LISTALL message format. This
+ * function may fail if another process is changing the DOI list at the same
+ * time. Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ struct cipso_v4_doi *iter;
+ u32 doi_cnt = 0;
+ ssize_t buf_len;
+
+ /* XXX - In both cases, this is kinda ugly as we have to go through
+ the list once to determine how large of a buffer we need,
+ drop the locks, allocate the buffer, grab the locks, and
+ finally fill the buffer. The problem is that there is that
+ open window where the table could grow and we will end up
+ short on space. */
+
+ buf_len = NETLBL_LEN_U32;
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+ if (iter->valid) {
+ doi_cnt += 1;
+ buf_len += 2 * NETLBL_LEN_U32;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+ netlbl_putinc_u32(&buf, doi_cnt, &buf_len);
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+ if (iter->valid) {
+ if (buf_len < 2 * NETLBL_LEN_U32) {
+ rcu_read_unlock();
+ kfree(skb);
+ return NULL;
+ }
+ netlbl_putinc_u32(&buf, iter->doi, &buf_len);
+ netlbl_putinc_u32(&buf, iter->type, &buf_len);
+ }
+ rcu_read_unlock();
+
+ return skb;
+}
+
+/**
+ * cipso_v4_doi_dump - Dump a CIPSO DOI definition into a sk_buff
+ * @doi: the DOI value
+ * @headroom: the amount of headroom to allocate for the sk_buff
+ *
+ * Description:
+ * Lookup the DOI definition matching @doi and dump it's contents into a
+ * sk_buff. The returned sk_buff has room at the front of the sk_buff for
+ * @headroom bytes. See net/netlabel/netlabel_cipso_v4.h for the LIST message
+ * format. This function may fail if another process is changing the DOI list
+ * at the same time. Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *cipso_v4_doi_dump(const u32 doi, const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ struct cipso_v4_doi *iter;
+ u32 tag_cnt = 0;
+ u32 lvl_cnt = 0;
+ u32 cat_cnt = 0;
+ ssize_t buf_len;
+ ssize_t tmp;
+
+ /* XXX - In both cases, this is kinda ugly as we have to go through
+ the list once to determine how large of a buffer we need,
+ drop the locks, allocate the buffer, grab the locks, and
+ finally fill the buffer. The problem is that there is that
+ open window where the table could grow and we will end up
+ short on space. */
+
+ rcu_read_lock();
+ iter = cipso_v4_doi_getdef(doi);
+ if (iter == NULL) {
+ rcu_read_unlock();
+ return NULL;
+ }
+ buf_len = NETLBL_LEN_U32;
+ switch (iter->type) {
+ case CIPSO_V4_MAP_PASS:
+ buf_len += NETLBL_LEN_U32;
+ while(tag_cnt < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tag_cnt] != CIPSO_V4_TAG_INVALID) {
+ tag_cnt += 1;
+ buf_len += NETLBL_LEN_U8;
+ }
+ break;
+ case CIPSO_V4_MAP_STD:
+ buf_len += 3 * NETLBL_LEN_U32;
+ while (tag_cnt < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tag_cnt] != CIPSO_V4_TAG_INVALID) {
+ tag_cnt += 1;
+ buf_len += NETLBL_LEN_U8;
+ }
+ for (tmp = 0; tmp < iter->map.std->lvl.local_size; tmp++)
+ if (iter->map.std->lvl.local[tmp] !=
+ CIPSO_V4_INV_LVL) {
+ lvl_cnt += 1;
+ buf_len += NETLBL_LEN_U32 + NETLBL_LEN_U8;
+ }
+ for (tmp = 0; tmp < iter->map.std->cat.local_size; tmp++)
+ if (iter->map.std->cat.local[tmp] !=
+ CIPSO_V4_INV_CAT) {
+ cat_cnt += 1;
+ buf_len += NETLBL_LEN_U32 + NETLBL_LEN_U16;
+ }
+ break;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+ netlbl_putinc_u32(&buf, iter->type, &buf_len);
+
+ rcu_read_lock();
+ if (iter != cipso_v4_doi_getdef(doi))
+ goto doi_dump_failure;
+ switch (iter->type) {
+ case CIPSO_V4_MAP_PASS:
+ netlbl_putinc_u32(&buf, tag_cnt, &buf_len);
+ for (tmp = 0;
+ tmp < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tmp] != CIPSO_V4_TAG_INVALID;
+ tmp++) {
+ if (buf_len < NETLBL_LEN_U8)
+ goto doi_dump_failure;
+ netlbl_putinc_u8(&buf, iter->tags[tmp], &buf_len);
+ }
+ break;
+ case CIPSO_V4_MAP_STD:
+ netlbl_putinc_u32(&buf, tag_cnt, &buf_len);
+ netlbl_putinc_u32(&buf, lvl_cnt, &buf_len);
+ netlbl_putinc_u32(&buf, cat_cnt, &buf_len);
+ for (tmp = 0;
+ tmp < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tmp] != CIPSO_V4_TAG_INVALID;
+ tmp++) {
+ if (buf_len < NETLBL_LEN_U8)
+ goto doi_dump_failure;
+ netlbl_putinc_u8(&buf, iter->tags[tmp], &buf_len);
+ }
+ for (tmp = 0; tmp < iter->map.std->lvl.local_size; tmp++)
+ if (iter->map.std->lvl.local[tmp] !=
+ CIPSO_V4_INV_LVL) {
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto doi_dump_failure;
+ netlbl_putinc_u32(&buf, tmp, &buf_len);
+ netlbl_putinc_u8(&buf,
+ iter->map.std->lvl.local[tmp],
+ &buf_len);
+ }
+ for (tmp = 0; tmp < iter->map.std->cat.local_size; tmp++)
+ if (iter->map.std->cat.local[tmp] !=
+ CIPSO_V4_INV_CAT) {
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto doi_dump_failure;
+ netlbl_putinc_u32(&buf, tmp, &buf_len);
+ netlbl_putinc_u16(&buf,
+ iter->map.std->cat.local[tmp],
+ &buf_len);
+ }
+ break;
+ }
+ rcu_read_unlock();
+
+ return skb;
+
+doi_dump_failure:
+ rcu_read_unlock();
+ kfree(skb);
+ return NULL;
+}
+
+/**
+ * cipso_v4_doi_domhsh_add - Adds a domain entry to a DOI definition
+ * @doi_def: the DOI definition
+ * @domain: the domain to add
+ *
+ * Description:
+ * Adds the @domain to the the DOI specified by @doi_def, this function
+ * should only be called by external functions (i.e. NetLabel). This function
+ * does allocate memory. Returns zero on success, negative values on failure.
+ *
+ */
+int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def, const char *domain)
+{
+ struct cipso_v4_domhsh_entry *iter;
+ struct cipso_v4_domhsh_entry *new_dom;
+
+ new_dom = kzalloc(sizeof(*new_dom), GFP_KERNEL);
+ if (new_dom == NULL)
+ return -ENOMEM;
+ if (domain) {
+ new_dom->domain = kstrdup(domain, GFP_KERNEL);
+ if (new_dom->domain == NULL) {
+ kfree(new_dom);
+ return -ENOMEM;
+ }
+ }
+ new_dom->valid = 1;
+ INIT_RCU_HEAD(&new_dom->rcu);
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &doi_def->dom_list, list)
+ if (iter->valid &&
+ ((domain != NULL && iter->domain != NULL &&
+ strcmp(iter->domain, domain) == 0) ||
+ (domain == NULL && iter->domain == NULL))) {
+ rcu_read_unlock();
+ kfree(new_dom->domain);
+ kfree(new_dom);
+ return -EEXIST;
+ }
+ spin_lock(&cipso_v4_doi_list_lock);
+ list_add_tail_rcu(&new_dom->list, &doi_def->dom_list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * cipso_v4_doi_domhsh_remove - Removes a domain entry from a DOI definition
+ * @doi_def: the DOI definition
+ * @domain: the domain to remove
+ *
+ * Description:
+ * Removes the @domain from the DOI specified by @doi_def, this function
+ * should only be called by external functions (i.e. NetLabel). Returns zero
+ * on success and negative values on error.
+ *
+ */
+int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+ const char *domain)
+{
+ struct cipso_v4_domhsh_entry *iter;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &doi_def->dom_list, list)
+ if (iter->valid &&
+ ((domain != NULL && iter->domain != NULL &&
+ strcmp(iter->domain, domain) == 0) ||
+ (domain == NULL && iter->domain == NULL))) {
+ spin_lock(&cipso_v4_doi_list_lock);
+ iter->valid = 0;
+ list_del_rcu(&iter->list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ rcu_read_unlock();
+ call_rcu(&iter->rcu, cipso_v4_doi_domhsh_free);
+
+ return 0;
+ }
+ rcu_read_unlock();
+
+ return -ENOENT;
+}
+
+/*
+ * Label Mapping Functions
+ */
+
+/**
+ * cipso_v4_map_lvl_valid - Checks to see if the given level is understood
+ * @doi_def: the DOI definition
+ * @level: the level to check
+ *
+ * Description:
+ * Checks the given level against the given DOI definition and returns a
+ * negative value if the level does not have a valid mapping and a zero value
+ * if the level is defined by the DOI.
+ *
+ */
+static int cipso_v4_map_lvl_valid(const struct cipso_v4_doi *doi_def,
+ const u8 level)
+{
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ if (doi_def->map.std->lvl.cipso[level] < CIPSO_V4_INV_LVL)
+ return 0;
+ break;
+ }
+
+ return -EFAULT;
+}
+
+/**
+ * cipso_v4_map_lvl_hton - Perform a level mapping from the host to the network
+ * @doi_def: the DOI definition
+ * @host_lvl: the host MLS level
+ * @net_lvl: the network/CIPSO MLS level
+ *
+ * Description:
+ * Perform a label mapping to translate a local MLS level to the correct
+ * CIPSO level using the given DOI definition. Returns zero on success,
+ * negative values otherwise.
+ *
+ */
+static int cipso_v4_map_lvl_hton(const struct cipso_v4_doi *doi_def,
+ const u32 host_lvl,
+ u32 *net_lvl)
+{
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ *net_lvl = host_lvl;
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ if (host_lvl < doi_def->map.std->lvl.local_size) {
+ *net_lvl = doi_def->map.std->lvl.local[host_lvl];
+ return 0;
+ }
+ break;
+ }
+
+ return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_lvl_ntoh - Perform a level mapping from the network to the host
+ * @doi_def: the DOI definition
+ * @net_lvl: the network/CIPSO MLS level
+ * @host_lvl: the host MLS level
+ *
+ * Description:
+ * Perform a label mapping to translate a CIPSO level to the correct local MLS
+ * level using the given DOI definition. Returns zero on success, negative
+ * values otherwise.
+ *
+ */
+static int cipso_v4_map_lvl_ntoh(const struct cipso_v4_doi *doi_def,
+ const u32 net_lvl,
+ u32 *host_lvl)
+{
+ struct cipso_v4_std_map_tbl *map_tbl;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ *host_lvl = net_lvl;
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ map_tbl = doi_def->map.std;
+ if (net_lvl < map_tbl->lvl.cipso_size &&
+ map_tbl->lvl.cipso[net_lvl] < CIPSO_V4_INV_LVL) {
+ *host_lvl = doi_def->map.std->lvl.cipso[net_lvl];
+ return 0;
+ }
+ break;
+ }
+
+ return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_valid - Checks to see if the category bitmap is valid
+ * @doi_def: the DOI definition
+ * @bitmap: category bitmap
+ * @bitmap_len: bitmap length in bytes
+ *
+ * Description:
+ * Checks the given category bitmap against the given DOI definition and
+ * returns a negative value if any of the categories in the bitmap do not have
+ * a valid mapping and a zero value if all of the categories are valid.
+ *
+ */
+static int cipso_v4_map_cat_rbm_valid(const struct cipso_v4_doi *doi_def,
+ const unsigned char *bitmap,
+ const u32 bitmap_len)
+{
+ int cat = -1;
+ u32 bitmap_len_bits = bitmap_len * 8;
+ u32 cipso_cat_size = doi_def->map.std->cat.cipso_size;
+ u32 *cipso_array = doi_def->map.std->cat.cipso;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ for (;;) {
+ cat = cipso_v4_bitmap_walk(bitmap,
+ bitmap_len_bits,
+ cat + 1,
+ 1);
+ if (cat < 0)
+ break;
+ if (cat >= cipso_cat_size ||
+ cipso_array[cat] >= CIPSO_V4_INV_CAT)
+ return -EFAULT;
+ }
+
+ if (cat == -1)
+ return 0;
+ break;
+ }
+
+ return -EFAULT;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_hton - Perform a category mapping from host to network
+ * @doi_def: the DOI definition
+ * @host_cat: the category bitmap in host format
+ * @host_cat_len: the length of the host's category bitmap in bytes
+ * @net_cat: the zero'd out category bitmap in network/CIPSO format
+ * @net_cat_len: the length of the CIPSO bitmap in bytes
+ *
+ * Description:
+ * Perform a label mapping to translate a local MLS category bitmap to the
+ * correct CIPSO bitmap using the given DOI definition. Returns the minimum
+ * size in bytes of the network bitmap on success, negative values otherwise.
+ *
+ */
+static int cipso_v4_map_cat_rbm_hton(const struct cipso_v4_doi *doi_def,
+ const unsigned char *host_cat,
+ const u32 host_cat_len,
+ unsigned char *net_cat,
+ const u32 net_cat_len)
+{
+ int host_spot = -1;
+ u32 net_spot;
+ u32 net_spot_max = 0;
+ u32 host_clen_bits = host_cat_len * 8;
+ u32 net_clen_bits = net_cat_len * 8;
+ u32 host_cat_size = doi_def->map.std->cat.local_size;
+ u32 *host_cat_array = doi_def->map.std->cat.local;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ net_spot_max = host_cat_len - 1;
+ while (net_spot_max > 0 && host_cat[net_spot_max] == 0)
+ net_spot_max--;
+ if (net_spot_max > net_cat_len)
+ return -EINVAL;
+ memcpy(net_cat, host_cat, net_spot_max);
+ return net_spot_max;
+ case CIPSO_V4_MAP_STD:
+ for (;;) {
+ host_spot = cipso_v4_bitmap_walk(host_cat,
+ host_clen_bits,
+ host_spot + 1,
+ 1);
+ if (host_spot < 0)
+ break;
+ if (host_spot >= host_cat_size)
+ return -EPERM;
+
+ net_spot = host_cat_array[host_spot];
+ if (net_spot >= net_clen_bits)
+ return -ENOSPC;
+ cipso_v4_bitmap_setbit(net_cat, net_spot, 1);
+
+ if (net_spot > net_spot_max)
+ net_spot_max = net_spot;
+ }
+
+ if (host_spot == -2)
+ return -EFAULT;
+
+ if (++net_spot_max % 8)
+ return net_spot_max / 8 + 1;
+ return net_spot_max / 8;
+ }
+
+ return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_ntoh - Perform a category mapping from network to host
+ * @doi_def: the DOI definition
+ * @net_cat: the category bitmap in network/CIPSO format
+ * @net_cat_len: the length of the CIPSO bitmap in bytes
+ * @host_cat: the zero'd out category bitmap in host format
+ * @host_cat_len: the length of the host's category bitmap in bytes
+ *
+ * Description:
+ * Perform a label mapping to translate a CIPSO bitmap to the correct local
+ * MLS category bitmap using the given DOI definition. Returns the minimum
+ * size in bytes of the host bitmap on success, negative values otherwise.
+ *
+ */
+static int cipso_v4_map_cat_rbm_ntoh(const struct cipso_v4_doi *doi_def,
+ const unsigned char *net_cat,
+ const u32 net_cat_len,
+ unsigned char *host_cat,
+ const u32 host_cat_len)
+{
+ u32 host_spot;
+ u32 host_spot_max = 0;
+ int net_spot = -1;
+ u32 net_clen_bits = net_cat_len * 8;
+ u32 host_clen_bits = host_cat_len * 8;
+ u32 net_cat_size = doi_def->map.std->cat.cipso_size;
+ u32 *net_cat_array = doi_def->map.std->cat.cipso;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ if (net_cat_len > host_cat_len)
+ return -EINVAL;
+ memcpy(host_cat, net_cat, net_cat_len);
+ return net_cat_len;
+ case CIPSO_V4_MAP_STD:
+ for (;;) {
+ net_spot = cipso_v4_bitmap_walk(net_cat,
+ net_clen_bits,
+ net_spot + 1,
+ 1);
+ if (net_spot < 0)
+ break;
+ if (net_spot >= net_cat_size ||
+ net_cat_array[net_spot] >= CIPSO_V4_INV_CAT)
+ return -EPERM;
+
+ host_spot = net_cat_array[net_spot];
+ if (host_spot >= host_clen_bits)
+ return -ENOSPC;
+ cipso_v4_bitmap_setbit(host_cat, host_spot, 1);
+
+ if (host_spot > host_spot_max)
+ host_spot_max = host_spot;
+ }
+
+ if (net_spot == -2)
+ return -EFAULT;
+
+ if (++host_spot_max % 8)
+ return host_spot_max / 8 + 1;
+ return host_spot_max / 8;
+ }
+
+ return -EINVAL;
+}
+
+/*
+ * Protocol Handling Functions
+ */
+
+#define CIPSO_V4_HDR_LEN 6
+
+/**
+ * cipso_v4_gentag_hdr - Generate a CIPSO option header
+ * @doi_def: the DOI definition
+ * @len: the total tag length in bytes
+ * @buf: the CIPSO option buffer
+ *
+ * Description:
+ * Write a CIPSO header into the beginning of @buffer. Return zero on success,
+ * negative values on failure.
+ *
+ */
+static int cipso_v4_gentag_hdr(const struct cipso_v4_doi *doi_def,
+ const u32 len,
+ unsigned char *buf)
+{
+ if (CIPSO_V4_HDR_LEN + len > 40)
+ return -ENOSPC;
+
+ buf[0] = IPOPT_CIPSO;
+ buf[1] = CIPSO_V4_HDR_LEN + len;
+ *(u32 *)&buf[2] = htonl(doi_def->doi);
+
+ return 0;
+}
+
+#define CIPSO_V4_TAG1_CAT_LEN 30
+
+/**
+ * cipso_v4_gentag_rbm - Generate a CIPSO restricted bitmap tag (type #1)
+ * @doi_def: the DOI definition
+ * @secattr: the security attributes
+ * @buffer: the option buffer
+ * @buffer_len: length of buffer in bytes
+ *
+ * Description:
+ * Generate a CIPSO option using the restricted bitmap tag, tag type #1. The
+ * actual buffer length may be larger than the indicated size due to
+ * translation between host and network category bitmaps. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int cipso_v4_gentag_rbm(const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr,
+ unsigned char **buffer,
+ u32 *buffer_len)
+{
+ int ret_val = -EPERM;
+ unsigned char *buf = NULL;
+ u32 buf_len;
+ u32 level;
+
+ if (secattr->mls_cat) {
+ buf = kzalloc(CIPSO_V4_HDR_LEN + 4 + CIPSO_V4_TAG1_CAT_LEN,
+ GFP_ATOMIC);
+ if (buf == NULL)
+ return -ENOMEM;
+
+ ret_val = cipso_v4_map_cat_rbm_hton(doi_def,
+ secattr->mls_cat,
+ secattr->mls_cat_len,
+ &buf[CIPSO_V4_HDR_LEN + 4],
+ CIPSO_V4_TAG1_CAT_LEN);
+ if (ret_val < 0)
+ goto gentag_failure;
+
+ /* XXX - this will send packets using the "optimized" format
+ when possibile as specified in section 3.4.2.6 of the
+ CIPSO draft */
+ if (cipso_v4_rbm_optfmt && (ret_val > 0 && ret_val < 10))
+ ret_val = 10;
+
+ buf_len = 4 + ret_val;
+ } else {
+ buf = kzalloc(CIPSO_V4_HDR_LEN + 4, GFP_ATOMIC);
+ if (buf == NULL)
+ return -ENOMEM;
+ buf_len = 4;
+ }
+
+ ret_val = cipso_v4_map_lvl_hton(doi_def, secattr->mls_lvl, &level);
+ if (ret_val != 0)
+ goto gentag_failure;
+
+ ret_val = cipso_v4_gentag_hdr(doi_def, buf_len, buf);
+ if (ret_val != 0)
+ goto gentag_failure;
+
+ buf[CIPSO_V4_HDR_LEN] = 0x01;
+ buf[CIPSO_V4_HDR_LEN + 1] = buf_len;
+ buf[CIPSO_V4_HDR_LEN + 3] = level;
+
+ *buffer = buf;
+ *buffer_len = CIPSO_V4_HDR_LEN + buf_len;
+
+ return 0;
+
+gentag_failure:
+ kfree(buf);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_parsetag_rbm - Parse a CIPSO restricted bitmap tag
+ * @doi_def: the DOI definition
+ * @tag: the CIPSO tag
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Parse a CIPSO restricted bitmap tag (tag type #1) and return the security
+ * attributes in @secattr. Return zero on success, negatives values on
+ * failure.
+ *
+ */
+static int cipso_v4_parsetag_rbm(const struct cipso_v4_doi *doi_def,
+ const unsigned char *tag,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+ u8 tag_len = tag[1];
+ u32 level;
+
+ ret_val = cipso_v4_map_lvl_ntoh(doi_def, tag[3], &level);
+ if (ret_val != 0)
+ return ret_val;
+ secattr->mls_lvl = level;
+ secattr->mls_lvl_vld = 1;
+
+ if (tag_len > 4) {
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ secattr->mls_cat_len = tag_len - 4;
+ break;
+ case CIPSO_V4_MAP_STD:
+ secattr->mls_cat_len =
+ doi_def->map.std->cat.local_size;
+ break;
+ }
+ secattr->mls_cat = kzalloc(secattr->mls_cat_len, GFP_ATOMIC);
+ if (secattr->mls_cat == NULL)
+ return -ENOMEM;
+
+ ret_val = cipso_v4_map_cat_rbm_ntoh(doi_def,
+ &tag[4],
+ tag_len - 4,
+ secattr->mls_cat,
+ secattr->mls_cat_len);
+ if (ret_val < 0) {
+ kfree(secattr->mls_cat);
+ return ret_val;
+ }
+ secattr->mls_cat_len = ret_val;
+ }
+
+ return 0;
+}
+
+/**
+ * cipso_v4_validate - Validate a CIPSO option
+ * @option: the start of the option, on error it is set to point to the error
+ *
+ * Description:
+ * This routine is called to validate a CIPSO option, it checks all of the
+ * fields to ensure that they are at least valid, see the draft snippet below
+ * for details. If the option is valid then a zero value is returned and
+ * the value of @option is unchanged. If the option is invalid then a
+ * non-zero value is returned and @option is adjusted to point to the
+ * offending portion of the option. From the IETF draft ...
+ *
+ * "If any field within the CIPSO options, such as the DOI identifier, is not
+ * recognized the IP datagram is discarded and an ICMP 'parameter problem'
+ * (type 12) is generated and returned. The ICMP code field is set to 'bad
+ * parameter' (code 0) and the pointer is set to the start of the CIPSO field
+ * that is unrecognized."
+ *
+ */
+int cipso_v4_validate(unsigned char **option)
+{
+ unsigned char *opt = *option;
+ unsigned char *tag;
+ unsigned char opt_iter;
+ unsigned char err_offset = 0;
+ unsigned char locked = 0;
+ u8 opt_len;
+ u8 tag_len;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 tag_iter;
+
+ /* XXX - caller already checks for length values that are too large */
+ opt_len = opt[1];
+ if (opt_len < 8) {
+ err_offset = 1;
+ goto validate_return;
+ }
+
+ rcu_read_lock();
+ locked = 1;
+ doi_def = cipso_v4_doi_getdef(ntohl(*((u32 *)&opt[2])));
+ if (doi_def == NULL) {
+ err_offset = 2;
+ goto validate_return;
+ }
+
+ opt_iter = 6;
+ tag = opt + opt_iter;
+ while (opt_iter < opt_len) {
+ for (tag_iter = 0; doi_def->tags[tag_iter] != tag[0];)
+ if (doi_def->tags[tag_iter] == CIPSO_V4_TAG_INVALID ||
+ ++tag_iter == CIPSO_V4_TAG_MAXCNT) {
+ err_offset = opt_iter;
+ goto validate_return;
+ }
+
+ tag_len = tag[1];
+ if (tag_len > (opt_len - opt_iter)) {
+ err_offset = opt_iter + 1;
+ goto validate_return;
+ }
+
+ switch (tag[0]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ if (tag_len < 4) {
+ err_offset = opt_iter + 1;
+ goto validate_return;
+ }
+
+ /* XXX - we are already going to do all the
+ verification necessary at the socket layer so from
+ our point of view it is safe to turn these checks
+ off (and less work), however, the CIPSO draft says
+ we should do all the CIPSO validations here but it
+ doesn't really specify _exactly_ what we need to
+ validate ... so, just make it a sysctl tunable */
+ if (cipso_v4_rbm_strictvalid) {
+ if (cipso_v4_map_lvl_valid(doi_def,
+ tag[3]) < 0) {
+ err_offset = opt_iter + 3;
+ goto validate_return;
+ }
+ if (tag_len > 4 &&
+ cipso_v4_map_cat_rbm_valid(doi_def,
+ &tag[4],
+ tag_len - 4) < 0) {
+ err_offset = opt_iter + 4;
+ goto validate_return;
+ }
+ }
+ break;
+ default:
+ err_offset = opt_iter;
+ goto validate_return;
+ }
+
+ tag += tag_len;
+ opt_iter += tag_len;
+ }
+
+validate_return:
+ if (locked)
+ rcu_read_unlock();
+ *option = opt + err_offset;
+ return err_offset;
+}
+
+/**
+ * cipso_v4_error - Send the correct reponse for a bad packet
+ * @skb: the packet
+ * @error: the error code
+ * @gateway: CIPSO gateway flag
+ *
+ * Description:
+ * Based on the error code given in @error, send an ICMP error message back to
+ * the originating host. From the IETF draft ...
+ *
+ * "If the contents of the CIPSO [option] are valid but the security label is
+ * outside of the configured host or port label range, the datagram is
+ * discarded and an ICMP 'destination unreachable' (type 3) is generated and
+ * returned. The code field of the ICMP is set to 'communication with
+ * destination network administratively prohibited' (code 9) or to
+ * 'communication with destination host administratively prohibited'
+ * (code 10). The value of the code is dependent on whether the originator
+ * of the ICMP message is acting as a CIPSO host or a CIPSO gateway. The
+ * recipient of the ICMP message MUST be able to handle either value. The
+ * same procedure is performed if a CIPSO [option] can not be added to an
+ * IP packet because it is too large to fit in the IP options area."
+ *
+ * "If the error is triggered by receipt of an ICMP message, the message is
+ * discarded and no response is permitted (consistent with general ICMP
+ * processing rules)."
+ *
+ */
+void cipso_v4_error(struct sk_buff *skb, const int error, const u32 gateway)
+{
+ if (skb->nh.iph->protocol == IPPROTO_ICMP || error != -EACCES)
+ return;
+
+ if (gateway)
+ icmp_send(skb, ICMP_DEST_UNREACH, ICMP_NET_ANO, 0);
+ else
+ icmp_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_ANO, 0);
+}
+
+/**
+ * cipso_v4_socket_setopt - Set the CIPSO option string on a socket
+ * @sock: the socket
+ * @opt: the option string
+ * @opt_len: the option string length
+ *
+ * Description:
+ * Attempt to add the given CIPSO option string to the socket. The caller is
+ * responsibile for locking and releasing the socket. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int cipso_v4_socket_setopt(struct socket *sock,
+ unsigned char *opt,
+ u32 opt_len)
+{
+ int ret_val;
+ struct ip_options *ip_opt = NULL;
+ struct sock *sk = sock->sk;
+ struct inet_sock *sk_inet;
+ struct inet_connection_sock *sk_conn;
+
+ ret_val = ip_options_get(&ip_opt, opt, opt_len);
+ if (ret_val != 0)
+ goto socket_setopt_return;
+ ip_opt->is_setbyuser = 0;
+ sk_inet = inet_sk(sk);
+ if (sk_inet->is_icsk) {
+ sk_conn = inet_csk(sk);
+ if (sk_inet->opt)
+ sk_conn->icsk_ext_hdr_len -= sk_inet->opt->optlen;
+ sk_conn->icsk_ext_hdr_len += ip_opt->optlen;
+ sk_conn->icsk_sync_mss(sk, sk_conn->icsk_pmtu_cookie);
+ }
+ ip_opt = xchg(&sk_inet->opt, ip_opt);
+
+ ret_val = 0;
+
+socket_setopt_return:
+ kfree(opt);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_socket_setattr - Add a CIPSO option to a socket
+ * @sock: the socket
+ * @doi_def: the CIPSO DOI to use
+ * @secattr: the specific security attributes of the socket
+ *
+ * Description:
+ * Set the CIPSO option on the given socket using the DOI definition and
+ * security attributes passed to the function. Returns zero on success and
+ * negative values on failure.
+ *
+ */
+int cipso_v4_socket_setattr(const struct socket *sock,
+ const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -EPERM;
+ u32 iter;
+ unsigned char *buf = NULL;
+ u32 buf_len = 0;
+ struct ip_options *opt = NULL;
+ struct sock *sk;
+ struct inet_sock *sk_inet;
+ struct inet_connection_sock *sk_conn;
+
+ /* In the case of sock_create_lite(), the sock->sk field is not
+ defined yet but it is not a problem as the only users of these
+ "lite" PF_INET sockets are functions which do an accept() call
+ afterwards so we will label the socket as part of the accept(). */
+ sk = sock->sk;
+ if (sk == NULL)
+ return 0;
+
+ /* XXX - this code assumes only one tag per CIPSO option which isnt
+ really a good assumption to make but since we only support the MAC
+ tags right now it is a safe assumption */
+ iter = 0;
+ do {
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ ret_val = cipso_v4_gentag_rbm(doi_def,
+ secattr,
+ &buf,
+ &buf_len);
+ break;
+ default:
+ ret_val = -EPERM;
+ goto socket_setattr_failure;
+ }
+
+ iter++;
+ } while (ret_val != 0 &&
+ iter < CIPSO_V4_TAG_MAXCNT &&
+ doi_def->tags[iter] != CIPSO_V4_TAG_INVALID);
+ if (ret_val != 0)
+ goto socket_setattr_failure;
+
+ ret_val = ip_options_get(&opt, buf, buf_len);
+ if (ret_val != 0)
+ goto socket_setattr_failure;
+ opt->is_setbyuser = 0;
+
+ kfree(buf);
+ buf = NULL;
+
+ lock_sock(sk);
+ sk_inet = inet_sk(sk);
+ if (sk_inet->is_icsk) {
+ sk_conn = inet_csk(sk);
+ if (sk_inet->opt)
+ sk_conn->icsk_ext_hdr_len -= sk_inet->opt->optlen;
+ sk_conn->icsk_ext_hdr_len += opt->optlen;
+ sk_conn->icsk_sync_mss(sk, sk_conn->icsk_pmtu_cookie);
+ }
+ opt = xchg(&sk_inet->opt, opt);
+ release_sock(sk);
+ kfree(opt);
+
+ return 0;
+
+socket_setattr_failure:
+ kfree(buf);
+ kfree(opt);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_socket_getopt - Get the CIPSO IP option string from the socket
+ * @sock: the socket
+ * @opt: the option string
+ * @opt_len: the option string length in bytes
+ *
+ * Description:
+ * If a CIPSO option is present in the socket's IP options then return a
+ * a pointer to the option string. The caller is responsibile for locking
+ * and releasing the socket. Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int cipso_v4_socket_getopt(const struct socket *sock,
+ unsigned char **opt,
+ u32 *opt_len)
+{
+ struct ip_options *ip_opt = inet_sk(sock->sk)->opt;
+ unsigned char *cipso_ptr;
+
+ if (ip_opt == NULL || ip_opt->cipso == 0)
+ return -ENOMSG;
+ cipso_ptr = ip_opt->__data + ip_opt->cipso - sizeof(struct iphdr);
+
+ *opt = cipso_ptr;
+ *opt_len = cipso_ptr[1];
+
+ return 0;
+}
+
+/**
+ * cipso_v4_socket_getattr - Get the security attributes from a socket
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Query @sock to see if there is a CIPSO option attached to the socket and if
+ * there is return the CIPSO security attributes in @secattr. Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -ENOMSG;
+ struct sock *sk;
+ struct inet_sock *sk_inet;
+ unsigned char *cipso_ptr;
+ u32 doi;
+ struct cipso_v4_doi *doi_def;
+
+ sk = sock->sk;
+ lock_sock(sk);
+ sk_inet = inet_sk(sk);
+ if (sk_inet->opt == NULL || sk_inet->opt->cipso == 0)
+ goto socket_getattr_return;
+ cipso_ptr = sk_inet->opt->__data + sk_inet->opt->cipso -
+ sizeof(struct iphdr);
+ ret_val = cipso_v4_cache_check(cipso_ptr, cipso_ptr[1], secattr);
+ if (ret_val == 0)
+ goto socket_getattr_return;
+
+ doi = ntohl(*(u32 *)&cipso_ptr[2]);
+ rcu_read_lock();
+ doi_def = cipso_v4_doi_getdef(doi);
+ if (doi_def == NULL) {
+ rcu_read_unlock();
+ goto socket_getattr_return;
+ }
+ switch (cipso_ptr[6]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ ret_val = cipso_v4_parsetag_rbm(doi_def,
+ &cipso_ptr[6],
+ secattr);
+ break;
+ }
+ rcu_read_unlock();
+
+socket_getattr_return:
+ release_sock(sk);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_skbuff_getattr - Get the security attributes from the CIPSO option
+ * @skb: the packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Parse the given packet's CIPSO option and return the security attributes.
+ * Returns zero on success and negative values on failure.
+ *
+ */
+int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -ENOMSG;
+ unsigned char *cipso_ptr;
+ u32 doi;
+ struct cipso_v4_doi *doi_def;
+
+ if (!CIPSO_V4_OPTEXIST(skb))
+ return -ENOMSG;
+ cipso_ptr = CIPSO_V4_OPTPTR(skb);
+ if (cipso_v4_cache_check(cipso_ptr, cipso_ptr[1], secattr) == 0)
+ return 0;
+
+ doi = ntohl(*(u32 *)&cipso_ptr[2]);
+ rcu_read_lock();
+ doi_def = cipso_v4_doi_getdef(doi);
+ if (doi_def == NULL)
+ goto skbuff_getattr_return;
+ switch (cipso_ptr[6]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ ret_val = cipso_v4_parsetag_rbm(doi_def,
+ &cipso_ptr[6],
+ secattr);
+ break;
+ }
+
+skbuff_getattr_return:
+ rcu_read_unlock();
+ return ret_val;
+}
+
+/*
+ * Setup Functions
+ */
+
+/**
+ * cipso_v4_init - Initialize the CIPSO module
+ *
+ * Description:
+ * Initialize the CIPSO module and prepare it for use. Returns zero on success
+ * and negative values on failure.
+ *
+ */
+static int __init cipso_v4_init(void)
+{
+ return cipso_v4_cache_init(CIPSO_V4_CACHE_BUCKETS);
+}
+
+/**
+ * cipso_v4_exit - Cleanup after the CIPSO module
+ *
+ * Description:
+ * Perform any necessary cleanup tasks before the CIPSO module goes away.
+ *
+ */
+static void __exit cipso_v4_exit(void)
+{
+ cipso_v4_cache_destroy();
+}
+
+subsys_initcall(cipso_v4_init);
Index: linux-2.6.18-rc1/net/ipv4/sysctl_net_ipv4.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/sysctl_net_ipv4.c
+++ linux-2.6.18-rc1/net/ipv4/sysctl_net_ipv4.c
@@ -17,6 +17,7 @@
#include <net/ip.h>
#include <net/route.h>
#include <net/tcp.h>
+#include <net/cipso_ipv4.h>
/* From af_inet.c */
extern int sysctl_ip_nonlocal_bind;
@@ -697,6 +698,40 @@ ctl_table ipv4_table[] = {
.mode = 0644,
.proc_handler = &proc_dointvec
},
+#ifdef CONFIG_NETLABEL
+ {
+ .ctl_name = NET_CIPSOV4_CACHE_ENABLE,
+ .procname = "cipso_cache_enable",
+ .data = &cipso_v4_cache_enabled,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
+ .ctl_name = NET_CIPSOV4_CACHE_BUCKET_SIZE,
+ .procname = "cipso_cache_bucket_size",
+ .data = &cipso_v4_cache_bucketsize,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
+ .ctl_name = NET_CIPSOV4_RBM_OPTFMT,
+ .procname = "cipso_rbm_optfmt",
+ .data = &cipso_v4_rbm_optfmt,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
+ .ctl_name = NET_CIPSOV4_RBM_STRICTVALID,
+ .procname = "cipso_rbm_strictvalid",
+ .data = &cipso_v4_rbm_strictvalid,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+#endif /* CONFIG_NETLABEL */
{ .ctl_name = 0 }
};
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Add support for the Commercial IP Security Option (CIPSO) to the IPv4 network
stack. CIPSO has become a de-facto standard for trusted/labeled networking
amongst existing Trusted Operating Systems such as Trusted Solaris, HP-UX CMW,
etc. This implementation is designed to be used with the NetLabel subsystem
to provide explicit packet labeling to LSM developers.
The CIPSO/IPv4 packet labeling works by the LSM calling a NetLabel API function
which attaches a CIPSO label (IPv4 option) to a given socket; this in turn
attaches the CIPSO label to every packet leaving the socket without any extra
processing on the outbound side. On the inbound side the individual packet's
sk_buff is examined through a call to a NetLabel API function to determine if a
CIPSO/IPv4 label is present and if so the security attributes of the CIPSO
label are returned to the caller of the NetLabel API function.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
include/linux/sysctl.h | 4
net/ipv4/Makefile | 1
net/ipv4/cipso_ipv4.c | 1708 +++++++++++++++++++++++++++++++++++++++++++++
net/ipv4/sysctl_net_ipv4.c | 35
4 files changed, 1748 insertions(+)
Index: linux-2.6.18-rc1/include/linux/sysctl.h
===================================================================
--- linux-2.6.18-rc1.orig/include/linux/sysctl.h
+++ linux-2.6.18-rc1/include/linux/sysctl.h
@@ -411,6 +411,10 @@ enum
NET_IPV4_TCP_WORKAROUND_SIGNED_WINDOWS=115,
NET_TCP_DMA_COPYBREAK=116,
NET_TCP_SLOW_START_AFTER_IDLE=117,
+ NET_CIPSOV4_CACHE_ENABLE=118,
+ NET_CIPSOV4_CACHE_BUCKET_SIZE=119,
+ NET_CIPSOV4_RBM_OPTFMT=120,
+ NET_CIPSOV4_RBM_STRICTVALID=121,
};
enum {
Index: linux-2.6.18-rc1/net/ipv4/Makefile
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/Makefile
+++ linux-2.6.18-rc1/net/ipv4/Makefile
@@ -48,6 +48,7 @@ obj-$(CONFIG_TCP_CONG_VENO) += tcp_veno.
obj-$(CONFIG_TCP_CONG_SCALABLE) += tcp_scalable.o
obj-$(CONFIG_TCP_CONG_LP) += tcp_lp.o
obj-$(CONFIG_TCP_CONG_COMPOUND) += tcp_compound.o
+obj-$(CONFIG_NETLABEL) += cipso_ipv4.o
obj-$(CONFIG_XFRM) += xfrm4_policy.o xfrm4_state.o xfrm4_input.o \
xfrm4_output.o
Index: linux-2.6.18-rc1/net/ipv4/cipso_ipv4.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/ipv4/cipso_ipv4.c
@@ -0,0 +1,1708 @@
+/*
+ * CIPSO - Commercial IP Security Option
+ *
+ * This is an implementation of the CIPSO 2.2 protocol as specified in
+ * draft-ietf-cipso-ipsecurity-01.txt with additional tag types as found in
+ * FIPS-188, copies of both documents can be found in the Documentation
+ * directory. While CIPSO never became a full IETF RFC standard many vendors
+ * have chosen to adopt the protocol and over the years it has become a
+ * de-facto standard for labeled networking.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <linux/module.h>
+#include <linux/jhash.h>
+#include <net/ip.h>
+#include <net/icmp.h>
+#include <net/tcp.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+struct cipso_v4_domhsh_entry {
+ char *domain;
+ u32 valid;
+ struct list_head list;
+ struct rcu_head rcu;
+};
+
+/* List of available DOI definitions */
+/* XXX - Updates should be minimal so having a single lock for the
+ cipso_v4_doi_list and the cipso_v4_doi_list->dom_list should be
+ okay. */
+/* XXX - This currently assumes a minimal number of different DOIs in use,
+ if in practice there are a lot of different DOIs this list should
+ probably be turned into a hash table or something similar so we
+ can do quick lookups. */
+DEFINE_SPINLOCK(cipso_v4_doi_list_lock);
+static struct list_head cipso_v4_doi_list = LIST_HEAD_INIT(cipso_v4_doi_list);
+
+/* Label mapping cache */
+int cipso_v4_cache_enabled = 1;
+int cipso_v4_cache_bucketsize = 10;
+#define CIPSO_V4_CACHE_BUCKETBITS 7
+#define CIPSO_V4_CACHE_BUCKETS (1 << CIPSO_V4_CACHE_BUCKETBITS)
+#define CIPSO_V4_CACHE_REORDERLIMIT 10
+struct cipso_v4_map_cache_bkt {
+ spinlock_t lock;
+ u32 size;
+ struct list_head list;
+};
+struct cipso_v4_map_cache_entry {
+ u32 hash;
+ unsigned char *key;
+ u32 key_len;
+
+ struct netlbl_lsm_cache lsm_data;
+
+ u32 activity;
+ struct list_head list;
+};
+static u32 cipso_v4_cache_size = 0;
+static struct cipso_v4_map_cache_bkt *cipso_v4_cache = NULL;
+#define CIPSO_V4_CACHE_ENABLED (cipso_v4_cache_enabled && \
+ cipso_v4_cache_bucketsize > 0 && \
+ cipso_v4_cache_size > 0)
+
+/* Restricted bitmap (tag #1) flags */
+int cipso_v4_rbm_optfmt = 0;
+int cipso_v4_rbm_strictvalid = 1;
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
+ * @bitmap: the bitmap
+ * @bitmap_len: length in bits
+ * @offset: starting offset
+ * @state: if non-zero, look for a set (1) bit else look for a cleared (0) bit
+ *
+ * Description:
+ * Starting at @offset, walk the bitmap from left to right until either the
+ * desired bit is found or we reach the end. Return the bit offset, -1 if
+ * not found, or -2 if error.
+ */
+static int cipso_v4_bitmap_walk(const unsigned char *bitmap,
+ const u32 bitmap_len,
+ const u32 offset,
+ const u8 state)
+{
+ u32 bit_spot;
+ u32 byte_offset;
+ unsigned char bitmask;
+ unsigned char byte;
+
+ /* gcc always rounds to zero when doing integer division */
+ byte_offset = offset / 8;
+ byte = bitmap[byte_offset];
+ bit_spot = offset;
+ bitmask = 0x80 >> (offset % 8);
+
+ /* XXX - probably should use include/asm/bitops.h if we can */
+ while (bit_spot < bitmap_len) {
+ if ((state && (byte & bitmask) == bitmask) ||
+ (state == 0 && (byte & bitmask) == 0))
+ return bit_spot;
+
+ bit_spot++;
+ bitmask >>= 1;
+ if (bitmask == 0) {
+ byte = bitmap[++byte_offset];
+ bitmask = 0x80;
+ }
+ }
+
+ return -1;
+}
+
+/**
+ * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
+ * @bitmap: the bitmap
+ * @bit: the bit
+ * @state: if non-zero, set the bit (1) else clear the bit (0)
+ *
+ * Description:
+ * Set a single bit in the bitmask. Returns zero on success, negative values
+ * on error.
+ */
+static void cipso_v4_bitmap_setbit(unsigned char *bitmap,
+ const u32 bit,
+ const u8 state)
+{
+ u32 byte_spot;
+ u8 bitmask;
+
+ /* gcc always rounds to zero when doing integer division */
+ byte_spot = bit / 8;
+ bitmask = 0x80 >> (bit % 8);
+ if (state)
+ bitmap[byte_spot] |= bitmask;
+ else
+ bitmap[byte_spot] &= ~bitmask;
+}
+
+/**
+ * cipso_v4_doi_domhsh_free - Frees a domain list entry
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to a domain list entry can be released
+ * safely.
+ *
+ */
+static void cipso_v4_doi_domhsh_free(struct rcu_head *entry)
+{
+ struct cipso_v4_domhsh_entry *ptr;
+
+ ptr = container_of(entry, struct cipso_v4_domhsh_entry, rcu);
+ kfree(ptr->domain);
+ kfree(ptr);
+}
+
+/**
+ * cipso_v4_cache_entry_free - Frees a cache entry
+ * @entry: the entry to free
+ *
+ * Description:
+ * This function frees the memory associated with a cache entry.
+ *
+ */
+static void cipso_v4_cache_entry_free(struct cipso_v4_map_cache_entry *entry)
+{
+ if (entry->lsm_data.free)
+ entry->lsm_data.free(entry->lsm_data.data);
+ kfree(entry->key);
+ kfree(entry);
+}
+
+/**
+ * cipso_v4_map_cache_hash - Hashing function for the CIPSO cache
+ * @key: the hash key
+ * @key_len: the length of the key in bytes
+ *
+ * Description:
+ * The CIPSO tag hashing function. Returns a 32-bit hash value.
+ *
+ */
+static u32 cipso_v4_map_cache_hash(const unsigned char *key, const u32 key_len)
+{
+ return jhash(key, key_len, 0);
+}
+
+/*
+ * Label Mapping Cache Functions
+ */
+
+/**
+ * cipso_v4_cache_init - Initialize the CIPSO cache
+ * @bkt_size: the number of cache buckets
+ *
+ * Description:
+ * Initializes the CIPSO label mapping cache, this function should be called
+ * before any of the other functions defined in this file. Returns zero on
+ * success, negative values on error.
+ *
+ */
+static int cipso_v4_cache_init(const u32 bkt_size)
+{
+ struct cipso_v4_map_cache_bkt *cache;
+ u32 iter;
+
+ if (bkt_size == 0)
+ return -EINVAL;
+
+ cache = kcalloc(bkt_size,
+ sizeof(struct cipso_v4_map_cache_bkt), GFP_KERNEL);
+ if (cache == NULL)
+ return -ENOMEM;
+
+ for (iter = 0; iter < bkt_size; iter++) {
+ cache[iter].lock = SPIN_LOCK_UNLOCKED;
+ cache[iter].size = 0;
+ INIT_LIST_HEAD(&cache[iter].list);
+ }
+ cipso_v4_cache = cache;
+ cipso_v4_cache_size = bkt_size;
+
+ return 0;
+}
+
+/**
+ * cipso_v4_cache_destroy - Destroy the CIPSO cache
+ *
+ * Description:
+ * Clears the CIPSO cache and frees all the memory. This function does not
+ * hold any locks and should only be called when the module is being unloaded.
+ *
+ */
+static int cipso_v4_cache_destroy(void)
+{
+ struct cipso_v4_map_cache_bkt *cache;
+ struct cipso_v4_map_cache_entry *entry;
+ u32 cache_size;
+ u32 iter;
+
+ cache = cipso_v4_cache;
+ cache_size = cipso_v4_cache_size;
+ cipso_v4_cache_size = 0;
+ cipso_v4_cache = NULL;
+
+ for (iter = 0; iter < cache_size; iter++)
+ list_for_each_entry(entry, &cache[iter].list, list) {
+ list_del(&entry->list);
+ cipso_v4_cache_entry_free(entry);
+ }
+
+ return 0;
+}
+
+/**
+ * cipso_v4_cache_invalidate - Invalidates the current CIPSO cache
+ *
+ * Description:
+ * Invalidates and frees any entries in the CIPSO cache. Returns zero on
+ * success and negative values on failure.
+ *
+ */
+void cipso_v4_cache_invalidate(void)
+{
+ struct cipso_v4_map_cache_entry *entry, *tmp_entry;
+ u32 iter;
+
+ if (cipso_v4_cache_size == 0)
+ return;
+
+ for (iter = 0; iter < cipso_v4_cache_size; iter++) {
+ spin_lock(&cipso_v4_cache[iter].lock);
+ list_for_each_entry_safe(entry,
+ tmp_entry,
+ &cipso_v4_cache[iter].list, list) {
+ list_del(&entry->list);
+ cipso_v4_cache_entry_free(entry);
+ }
+ cipso_v4_cache[iter].size = 0;
+ spin_unlock(&cipso_v4_cache[iter].lock);
+ }
+
+ return;
+}
+
+/**
+ * cipso_v4_cache_check - Check the CIPSO cache for a label mapping
+ * @key: the buffer to check
+ * @key_len: buffer length in bytes
+ * @secattr: the security attribute struct to use
+ *
+ * Description:
+ * This function checks the cache to see if a label mapping already exists for
+ * the given key. If there is a match then the cache is adjusted and the
+ * @secattr struct is populated with the correct LSM security attributes. The
+ * cache is adjusted in the following manner if the entry is not already the
+ * first in the cache bucket:
+ *
+ * 1. The cache entry's activity counter is incremented
+ * 2. The previous (higher ranking) entry's activity counter is decremented
+ * 3. If the difference between the two activity counters is geater than
+ * CIPSO_V4_CACHE_REORDERLIMIT the two entries are swapped
+ *
+ * Returns zero on success, -ENOENT for a cache miss, and other negative values
+ * on error.
+ *
+ */
+static int cipso_v4_cache_check(const unsigned char *key,
+ const u32 key_len,
+ struct netlbl_lsm_secattr *secattr)
+{
+ u32 bkt;
+ struct cipso_v4_map_cache_entry *entry;
+ struct cipso_v4_map_cache_entry *prev_entry = NULL;
+ u32 hash;
+
+ if (!CIPSO_V4_CACHE_ENABLED)
+ return -ENOENT;
+
+ hash = cipso_v4_map_cache_hash(key, key_len);
+ bkt = hash & (CIPSO_V4_CACHE_BUCKETBITS - 1);
+ spin_lock(&cipso_v4_cache[bkt].lock);
+ list_for_each_entry(entry, &cipso_v4_cache[bkt].list, list) {
+ if (entry->hash == hash &&
+ entry->key_len == key_len &&
+ memcmp(entry->key, key, key_len) == 0) {
+ entry->activity += 1;
+
+ if (prev_entry != NULL) {
+ if (prev_entry->activity > 0)
+ prev_entry->activity -= 1;
+
+ if (entry->activity > prev_entry->activity &&
+ entry->activity - prev_entry->activity >
+ CIPSO_V4_CACHE_REORDERLIMIT) {
+ __list_del(entry->list.prev,
+ entry->list.next);
+ __list_add(&entry->list,
+ prev_entry->list.prev,
+ &prev_entry->list);
+ }
+ }
+
+ secattr->cache.free = entry->lsm_data.free;
+ secattr->cache.data = entry->lsm_data.data;
+ spin_unlock(&cipso_v4_cache[bkt].lock);
+ return 0;
+ }
+ prev_entry = entry;
+ }
+ spin_unlock(&cipso_v4_cache[bkt].lock);
+
+ return -ENOENT;
+}
+
+/**
+ * cipso_v4_cache_add - Add an entry to the CIPSO cache
+ * @skb: the packet
+ * @secattr: the packet's security attributes
+ *
+ * Description:
+ * Add a new entry into the CIPSO label mapping cache. Add the new entry to
+ * head of the cache bucket's list, if the cache bucket is out of room remove
+ * the last entry in the list first. It is important to note that there is
+ * currently no checking for duplicate keys. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int cipso_v4_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -EPERM;
+ u32 bkt;
+ struct cipso_v4_map_cache_entry *entry = NULL;
+ struct cipso_v4_map_cache_entry *old_entry = NULL;
+ unsigned char *cipso_ptr;
+ u32 cipso_ptr_len;
+
+ if (!CIPSO_V4_CACHE_ENABLED)
+ return 0;
+
+ cipso_ptr = CIPSO_V4_OPTPTR(skb);
+ cipso_ptr_len = cipso_ptr[1];
+
+ entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
+ if (entry == NULL)
+ return -ENOMEM;
+ entry->key = kmalloc(cipso_ptr_len, GFP_ATOMIC);
+ if (entry->key == NULL) {
+ ret_val = -ENOMEM;
+ goto cache_add_failure;
+ }
+ memcpy(entry->key, cipso_ptr, cipso_ptr_len);
+ entry->key_len = cipso_ptr_len;
+ entry->hash = cipso_v4_map_cache_hash(cipso_ptr, cipso_ptr_len);
+ entry->lsm_data.free = secattr->cache.free;
+ entry->lsm_data.data = secattr->cache.data;
+
+ bkt = entry->hash & (CIPSO_V4_CACHE_BUCKETBITS - 1);
+ spin_lock(&cipso_v4_cache[bkt].lock);
+ if (cipso_v4_cache[bkt].size < cipso_v4_cache_bucketsize) {
+ list_add(&entry->list, &cipso_v4_cache[bkt].list);
+ cipso_v4_cache[bkt].size += 1;
+ } else {
+ old_entry = list_entry(cipso_v4_cache[bkt].list.prev,
+ struct cipso_v4_map_cache_entry, list);
+ list_del(&old_entry->list);
+ list_add(&entry->list, &cipso_v4_cache[bkt].list);
+ cipso_v4_cache_entry_free(old_entry);
+ }
+ spin_unlock(&cipso_v4_cache[bkt].lock);
+
+ return 0;
+
+cache_add_failure:
+ if (entry)
+ cipso_v4_cache_entry_free(entry);
+ return ret_val;
+}
+
+/*
+ * DOI List Functions
+ */
+
+/**
+ * cipso_v4_doi_search - Searches for a DOI definition
+ * @doi: the DOI to search for
+ *
+ * Description:
+ * Search the DOI definition list for a DOI definition with a DOI value that
+ * matches @doi. The caller is responsibile for calling rcu_read_[un]lock().
+ * Returns a pointer to the DOI definition on success and NULL on failure.
+ */
+static struct cipso_v4_doi *cipso_v4_doi_search(const u32 doi)
+{
+ struct cipso_v4_doi *iter;
+
+ list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+ if (iter->doi == doi && iter->valid)
+ return iter;
+ return NULL;
+}
+
+/**
+ * cipso_v4_doi_add - Add a new DOI to the CIPSO protocol engine
+ * @doi_def: the DOI structure
+ *
+ * Description:
+ * The caller defines a new DOI for use by the CIPSO engine and calls this
+ * function to add it to the list of acceptable domains. The caller must
+ * ensure that the mapping table specified in @doi_def->map meets all of the
+ * requirements of the mapping type (see cipso_ipv4.h for details). Returns
+ * zero on success and non-zero on failure.
+ *
+ */
+int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
+{
+ if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
+ return -EINVAL;
+
+ doi_def->valid = 1;
+ INIT_RCU_HEAD(&doi_def->rcu);
+ INIT_LIST_HEAD(&doi_def->dom_list);
+
+ rcu_read_lock();
+ if (cipso_v4_doi_search(doi_def->doi) != NULL) {
+ rcu_read_unlock();
+ return -EEXIST;
+ }
+ spin_lock(&cipso_v4_doi_list_lock);
+ list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * cipso_v4_doi_remove - Remove an existing DOI from the CIPSO protocol engine
+ * @doi: the DOI value
+ * @callback: the DOI cleanup/free callback
+ *
+ * Description:
+ * Removes a DOI definition from the CIPSO engine, @callback is called to
+ * free any memory. The NetLabel routines will be called to release their own
+ * LSM domain mappings as well as our own domain list. Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_doi_remove(const u32 doi,
+ void (*callback) (struct rcu_head * head))
+{
+ struct cipso_v4_doi *doi_def;
+ struct cipso_v4_domhsh_entry *dom_iter;
+
+ rcu_read_lock();
+ doi_def = cipso_v4_doi_search(doi);
+ if (doi_def != NULL) {
+ spin_lock(&cipso_v4_doi_list_lock);
+ doi_def->valid = 0;
+ list_del_rcu(&doi_def->list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ list_for_each_entry_rcu(dom_iter, &doi_def->dom_list, list)
+ if (dom_iter->valid)
+ netlbl_domhsh_remove(dom_iter->domain);
+ cipso_v4_cache_invalidate();
+ rcu_read_unlock();
+
+ call_rcu(&doi_def->rcu, callback);
+ return 0;
+ }
+ rcu_read_unlock();
+
+ return -ENOENT;
+}
+
+/**
+ * cipso_v4_doi_getdef - Returns a pointer to a valid DOI definition
+ * @doi: the DOI value
+ *
+ * Description:
+ * Searches for a valid DOI definition and if one is found it is returned to
+ * the caller. Otherwise NULL is returned. The caller must ensure that
+ * rcu_read_lock() is held while accessing the returned definition.
+ *
+ */
+struct cipso_v4_doi *cipso_v4_doi_getdef(const u32 doi)
+{
+ return cipso_v4_doi_search(doi);
+}
+
+/**
+ * cipso_v4_doi_dump_all - Dump all the CIPSO DOI definitions into a sk_buff
+ * @headroom: the amount of headroom to allocate for the sk_buff
+ *
+ * Description:
+ * Dump a list of all the configured DOI values into a sk_buff. The returned
+ * sk_buff has room at the front of the sk_buff for @headroom bytes. See
+ * net/netlabel/netlabel_cipso_v4.h for the LISTALL message format. This
+ * function may fail if another process is changing the DOI list at the same
+ * time. Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ struct cipso_v4_doi *iter;
+ u32 doi_cnt = 0;
+ ssize_t buf_len;
+
+ /* XXX - In both cases, this is kinda ugly as we have to go through
+ the list once to determine how large of a buffer we need,
+ drop the locks, allocate the buffer, grab the locks, and
+ finally fill the buffer. The problem is that there is that
+ open window where the table could grow and we will end up
+ short on space. */
+
+ buf_len = NETLBL_LEN_U32;
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+ if (iter->valid) {
+ doi_cnt += 1;
+ buf_len += 2 * NETLBL_LEN_U32;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+ netlbl_putinc_u32(&buf, doi_cnt, &buf_len);
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+ if (iter->valid) {
+ if (buf_len < 2 * NETLBL_LEN_U32) {
+ rcu_read_unlock();
+ kfree(skb);
+ return NULL;
+ }
+ netlbl_putinc_u32(&buf, iter->doi, &buf_len);
+ netlbl_putinc_u32(&buf, iter->type, &buf_len);
+ }
+ rcu_read_unlock();
+
+ return skb;
+}
+
+/**
+ * cipso_v4_doi_dump - Dump a CIPSO DOI definition into a sk_buff
+ * @doi: the DOI value
+ * @headroom: the amount of headroom to allocate for the sk_buff
+ *
+ * Description:
+ * Lookup the DOI definition matching @doi and dump it's contents into a
+ * sk_buff. The returned sk_buff has room at the front of the sk_buff for
+ * @headroom bytes. See net/netlabel/netlabel_cipso_v4.h for the LIST message
+ * format. This function may fail if another process is changing the DOI list
+ * at the same time. Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *cipso_v4_doi_dump(const u32 doi, const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ struct cipso_v4_doi *iter;
+ u32 tag_cnt = 0;
+ u32 lvl_cnt = 0;
+ u32 cat_cnt = 0;
+ ssize_t buf_len;
+ ssize_t tmp;
+
+ /* XXX - In both cases, this is kinda ugly as we have to go through
+ the list once to determine how large of a buffer we need,
+ drop the locks, allocate the buffer, grab the locks, and
+ finally fill the buffer. The problem is that there is that
+ open window where the table could grow and we will end up
+ short on space. */
+
+ rcu_read_lock();
+ iter = cipso_v4_doi_getdef(doi);
+ if (iter == NULL) {
+ rcu_read_unlock();
+ return NULL;
+ }
+ buf_len = NETLBL_LEN_U32;
+ switch (iter->type) {
+ case CIPSO_V4_MAP_PASS:
+ buf_len += NETLBL_LEN_U32;
+ while(tag_cnt < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tag_cnt] != CIPSO_V4_TAG_INVALID) {
+ tag_cnt += 1;
+ buf_len += NETLBL_LEN_U8;
+ }
+ break;
+ case CIPSO_V4_MAP_STD:
+ buf_len += 3 * NETLBL_LEN_U32;
+ while (tag_cnt < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tag_cnt] != CIPSO_V4_TAG_INVALID) {
+ tag_cnt += 1;
+ buf_len += NETLBL_LEN_U8;
+ }
+ for (tmp = 0; tmp < iter->map.std->lvl.local_size; tmp++)
+ if (iter->map.std->lvl.local[tmp] !=
+ CIPSO_V4_INV_LVL) {
+ lvl_cnt += 1;
+ buf_len += NETLBL_LEN_U32 + NETLBL_LEN_U8;
+ }
+ for (tmp = 0; tmp < iter->map.std->cat.local_size; tmp++)
+ if (iter->map.std->cat.local[tmp] !=
+ CIPSO_V4_INV_CAT) {
+ cat_cnt += 1;
+ buf_len += NETLBL_LEN_U32 + NETLBL_LEN_U16;
+ }
+ break;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+ netlbl_putinc_u32(&buf, iter->type, &buf_len);
+
+ rcu_read_lock();
+ if (iter != cipso_v4_doi_getdef(doi))
+ goto doi_dump_failure;
+ switch (iter->type) {
+ case CIPSO_V4_MAP_PASS:
+ netlbl_putinc_u32(&buf, tag_cnt, &buf_len);
+ for (tmp = 0;
+ tmp < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tmp] != CIPSO_V4_TAG_INVALID;
+ tmp++) {
+ if (buf_len < NETLBL_LEN_U8)
+ goto doi_dump_failure;
+ netlbl_putinc_u8(&buf, iter->tags[tmp], &buf_len);
+ }
+ break;
+ case CIPSO_V4_MAP_STD:
+ netlbl_putinc_u32(&buf, tag_cnt, &buf_len);
+ netlbl_putinc_u32(&buf, lvl_cnt, &buf_len);
+ netlbl_putinc_u32(&buf, cat_cnt, &buf_len);
+ for (tmp = 0;
+ tmp < CIPSO_V4_TAG_MAXCNT &&
+ iter->tags[tmp] != CIPSO_V4_TAG_INVALID;
+ tmp++) {
+ if (buf_len < NETLBL_LEN_U8)
+ goto doi_dump_failure;
+ netlbl_putinc_u8(&buf, iter->tags[tmp], &buf_len);
+ }
+ for (tmp = 0; tmp < iter->map.std->lvl.local_size; tmp++)
+ if (iter->map.std->lvl.local[tmp] !=
+ CIPSO_V4_INV_LVL) {
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto doi_dump_failure;
+ netlbl_putinc_u32(&buf, tmp, &buf_len);
+ netlbl_putinc_u8(&buf,
+ iter->map.std->lvl.local[tmp],
+ &buf_len);
+ }
+ for (tmp = 0; tmp < iter->map.std->cat.local_size; tmp++)
+ if (iter->map.std->cat.local[tmp] !=
+ CIPSO_V4_INV_CAT) {
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto doi_dump_failure;
+ netlbl_putinc_u32(&buf, tmp, &buf_len);
+ netlbl_putinc_u16(&buf,
+ iter->map.std->cat.local[tmp],
+ &buf_len);
+ }
+ break;
+ }
+ rcu_read_unlock();
+
+ return skb;
+
+doi_dump_failure:
+ rcu_read_unlock();
+ kfree(skb);
+ return NULL;
+}
+
+/**
+ * cipso_v4_doi_domhsh_add - Adds a domain entry to a DOI definition
+ * @doi_def: the DOI definition
+ * @domain: the domain to add
+ *
+ * Description:
+ * Adds the @domain to the the DOI specified by @doi_def, this function
+ * should only be called by external functions (i.e. NetLabel). This function
+ * does allocate memory. Returns zero on success, negative values on failure.
+ *
+ */
+int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def, const char *domain)
+{
+ struct cipso_v4_domhsh_entry *iter;
+ struct cipso_v4_domhsh_entry *new_dom;
+
+ new_dom = kzalloc(sizeof(*new_dom), GFP_KERNEL);
+ if (new_dom == NULL)
+ return -ENOMEM;
+ if (domain) {
+ new_dom->domain = kstrdup(domain, GFP_KERNEL);
+ if (new_dom->domain == NULL) {
+ kfree(new_dom);
+ return -ENOMEM;
+ }
+ }
+ new_dom->valid = 1;
+ INIT_RCU_HEAD(&new_dom->rcu);
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &doi_def->dom_list, list)
+ if (iter->valid &&
+ ((domain != NULL && iter->domain != NULL &&
+ strcmp(iter->domain, domain) == 0) ||
+ (domain == NULL && iter->domain == NULL))) {
+ rcu_read_unlock();
+ kfree(new_dom->domain);
+ kfree(new_dom);
+ return -EEXIST;
+ }
+ spin_lock(&cipso_v4_doi_list_lock);
+ list_add_tail_rcu(&new_dom->list, &doi_def->dom_list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * cipso_v4_doi_domhsh_remove - Removes a domain entry from a DOI definition
+ * @doi_def: the DOI definition
+ * @domain: the domain to remove
+ *
+ * Description:
+ * Removes the @domain from the DOI specified by @doi_def, this function
+ * should only be called by external functions (i.e. NetLabel). Returns zero
+ * on success and negative values on error.
+ *
+ */
+int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+ const char *domain)
+{
+ struct cipso_v4_domhsh_entry *iter;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &doi_def->dom_list, list)
+ if (iter->valid &&
+ ((domain != NULL && iter->domain != NULL &&
+ strcmp(iter->domain, domain) == 0) ||
+ (domain == NULL && iter->domain == NULL))) {
+ spin_lock(&cipso_v4_doi_list_lock);
+ iter->valid = 0;
+ list_del_rcu(&iter->list);
+ spin_unlock(&cipso_v4_doi_list_lock);
+ rcu_read_unlock();
+ call_rcu(&iter->rcu, cipso_v4_doi_domhsh_free);
+
+ return 0;
+ }
+ rcu_read_unlock();
+
+ return -ENOENT;
+}
+
+/*
+ * Label Mapping Functions
+ */
+
+/**
+ * cipso_v4_map_lvl_valid - Checks to see if the given level is understood
+ * @doi_def: the DOI definition
+ * @level: the level to check
+ *
+ * Description:
+ * Checks the given level against the given DOI definition and returns a
+ * negative value if the level does not have a valid mapping and a zero value
+ * if the level is defined by the DOI.
+ *
+ */
+static int cipso_v4_map_lvl_valid(const struct cipso_v4_doi *doi_def,
+ const u8 level)
+{
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ if (doi_def->map.std->lvl.cipso[level] < CIPSO_V4_INV_LVL)
+ return 0;
+ break;
+ }
+
+ return -EFAULT;
+}
+
+/**
+ * cipso_v4_map_lvl_hton - Perform a level mapping from the host to the network
+ * @doi_def: the DOI definition
+ * @host_lvl: the host MLS level
+ * @net_lvl: the network/CIPSO MLS level
+ *
+ * Description:
+ * Perform a label mapping to translate a local MLS level to the correct
+ * CIPSO level using the given DOI definition. Returns zero on success,
+ * negative values otherwise.
+ *
+ */
+static int cipso_v4_map_lvl_hton(const struct cipso_v4_doi *doi_def,
+ const u32 host_lvl,
+ u32 *net_lvl)
+{
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ *net_lvl = host_lvl;
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ if (host_lvl < doi_def->map.std->lvl.local_size) {
+ *net_lvl = doi_def->map.std->lvl.local[host_lvl];
+ return 0;
+ }
+ break;
+ }
+
+ return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_lvl_ntoh - Perform a level mapping from the network to the host
+ * @doi_def: the DOI definition
+ * @net_lvl: the network/CIPSO MLS level
+ * @host_lvl: the host MLS level
+ *
+ * Description:
+ * Perform a label mapping to translate a CIPSO level to the correct local MLS
+ * level using the given DOI definition. Returns zero on success, negative
+ * values otherwise.
+ *
+ */
+static int cipso_v4_map_lvl_ntoh(const struct cipso_v4_doi *doi_def,
+ const u32 net_lvl,
+ u32 *host_lvl)
+{
+ struct cipso_v4_std_map_tbl *map_tbl;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ *host_lvl = net_lvl;
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ map_tbl = doi_def->map.std;
+ if (net_lvl < map_tbl->lvl.cipso_size &&
+ map_tbl->lvl.cipso[net_lvl] < CIPSO_V4_INV_LVL) {
+ *host_lvl = doi_def->map.std->lvl.cipso[net_lvl];
+ return 0;
+ }
+ break;
+ }
+
+ return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_valid - Checks to see if the category bitmap is valid
+ * @doi_def: the DOI definition
+ * @bitmap: category bitmap
+ * @bitmap_len: bitmap length in bytes
+ *
+ * Description:
+ * Checks the given category bitmap against the given DOI definition and
+ * returns a negative value if any of the categories in the bitmap do not have
+ * a valid mapping and a zero value if all of the categories are valid.
+ *
+ */
+static int cipso_v4_map_cat_rbm_valid(const struct cipso_v4_doi *doi_def,
+ const unsigned char *bitmap,
+ const u32 bitmap_len)
+{
+ int cat = -1;
+ u32 bitmap_len_bits = bitmap_len * 8;
+ u32 cipso_cat_size = doi_def->map.std->cat.cipso_size;
+ u32 *cipso_array = doi_def->map.std->cat.cipso;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ return 0;
+ case CIPSO_V4_MAP_STD:
+ for (;;) {
+ cat = cipso_v4_bitmap_walk(bitmap,
+ bitmap_len_bits,
+ cat + 1,
+ 1);
+ if (cat < 0)
+ break;
+ if (cat >= cipso_cat_size ||
+ cipso_array[cat] >= CIPSO_V4_INV_CAT)
+ return -EFAULT;
+ }
+
+ if (cat == -1)
+ return 0;
+ break;
+ }
+
+ return -EFAULT;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_hton - Perform a category mapping from host to network
+ * @doi_def: the DOI definition
+ * @host_cat: the category bitmap in host format
+ * @host_cat_len: the length of the host's category bitmap in bytes
+ * @net_cat: the zero'd out category bitmap in network/CIPSO format
+ * @net_cat_len: the length of the CIPSO bitmap in bytes
+ *
+ * Description:
+ * Perform a label mapping to translate a local MLS category bitmap to the
+ * correct CIPSO bitmap using the given DOI definition. Returns the minimum
+ * size in bytes of the network bitmap on success, negative values otherwise.
+ *
+ */
+static int cipso_v4_map_cat_rbm_hton(const struct cipso_v4_doi *doi_def,
+ const unsigned char *host_cat,
+ const u32 host_cat_len,
+ unsigned char *net_cat,
+ const u32 net_cat_len)
+{
+ int host_spot = -1;
+ u32 net_spot;
+ u32 net_spot_max = 0;
+ u32 host_clen_bits = host_cat_len * 8;
+ u32 net_clen_bits = net_cat_len * 8;
+ u32 host_cat_size = doi_def->map.std->cat.local_size;
+ u32 *host_cat_array = doi_def->map.std->cat.local;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ net_spot_max = host_cat_len - 1;
+ while (net_spot_max > 0 && host_cat[net_spot_max] == 0)
+ net_spot_max--;
+ if (net_spot_max > net_cat_len)
+ return -EINVAL;
+ memcpy(net_cat, host_cat, net_spot_max);
+ return net_spot_max;
+ case CIPSO_V4_MAP_STD:
+ for (;;) {
+ host_spot = cipso_v4_bitmap_walk(host_cat,
+ host_clen_bits,
+ host_spot + 1,
+ 1);
+ if (host_spot < 0)
+ break;
+ if (host_spot >= host_cat_size)
+ return -EPERM;
+
+ net_spot = host_cat_array[host_spot];
+ if (net_spot >= net_clen_bits)
+ return -ENOSPC;
+ cipso_v4_bitmap_setbit(net_cat, net_spot, 1);
+
+ if (net_spot > net_spot_max)
+ net_spot_max = net_spot;
+ }
+
+ if (host_spot == -2)
+ return -EFAULT;
+
+ if (++net_spot_max % 8)
+ return net_spot_max / 8 + 1;
+ return net_spot_max / 8;
+ }
+
+ return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_ntoh - Perform a category mapping from network to host
+ * @doi_def: the DOI definition
+ * @net_cat: the category bitmap in network/CIPSO format
+ * @net_cat_len: the length of the CIPSO bitmap in bytes
+ * @host_cat: the zero'd out category bitmap in host format
+ * @host_cat_len: the length of the host's category bitmap in bytes
+ *
+ * Description:
+ * Perform a label mapping to translate a CIPSO bitmap to the correct local
+ * MLS category bitmap using the given DOI definition. Returns the minimum
+ * size in bytes of the host bitmap on success, negative values otherwise.
+ *
+ */
+static int cipso_v4_map_cat_rbm_ntoh(const struct cipso_v4_doi *doi_def,
+ const unsigned char *net_cat,
+ const u32 net_cat_len,
+ unsigned char *host_cat,
+ const u32 host_cat_len)
+{
+ u32 host_spot;
+ u32 host_spot_max = 0;
+ int net_spot = -1;
+ u32 net_clen_bits = net_cat_len * 8;
+ u32 host_clen_bits = host_cat_len * 8;
+ u32 net_cat_size = doi_def->map.std->cat.cipso_size;
+ u32 *net_cat_array = doi_def->map.std->cat.cipso;
+
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ if (net_cat_len > host_cat_len)
+ return -EINVAL;
+ memcpy(host_cat, net_cat, net_cat_len);
+ return net_cat_len;
+ case CIPSO_V4_MAP_STD:
+ for (;;) {
+ net_spot = cipso_v4_bitmap_walk(net_cat,
+ net_clen_bits,
+ net_spot + 1,
+ 1);
+ if (net_spot < 0)
+ break;
+ if (net_spot >= net_cat_size ||
+ net_cat_array[net_spot] >= CIPSO_V4_INV_CAT)
+ return -EPERM;
+
+ host_spot = net_cat_array[net_spot];
+ if (host_spot >= host_clen_bits)
+ return -ENOSPC;
+ cipso_v4_bitmap_setbit(host_cat, host_spot, 1);
+
+ if (host_spot > host_spot_max)
+ host_spot_max = host_spot;
+ }
+
+ if (net_spot == -2)
+ return -EFAULT;
+
+ if (++host_spot_max % 8)
+ return host_spot_max / 8 + 1;
+ return host_spot_max / 8;
+ }
+
+ return -EINVAL;
+}
+
+/*
+ * Protocol Handling Functions
+ */
+
+#define CIPSO_V4_HDR_LEN 6
+
+/**
+ * cipso_v4_gentag_hdr - Generate a CIPSO option header
+ * @doi_def: the DOI definition
+ * @len: the total tag length in bytes
+ * @buf: the CIPSO option buffer
+ *
+ * Description:
+ * Write a CIPSO header into the beginning of @buffer. Return zero on success,
+ * negative values on failure.
+ *
+ */
+static int cipso_v4_gentag_hdr(const struct cipso_v4_doi *doi_def,
+ const u32 len,
+ unsigned char *buf)
+{
+ if (CIPSO_V4_HDR_LEN + len > 40)
+ return -ENOSPC;
+
+ buf[0] = IPOPT_CIPSO;
+ buf[1] = CIPSO_V4_HDR_LEN + len;
+ *(u32 *)&buf[2] = htonl(doi_def->doi);
+
+ return 0;
+}
+
+#define CIPSO_V4_TAG1_CAT_LEN 30
+
+/**
+ * cipso_v4_gentag_rbm - Generate a CIPSO restricted bitmap tag (type #1)
+ * @doi_def: the DOI definition
+ * @secattr: the security attributes
+ * @buffer: the option buffer
+ * @buffer_len: length of buffer in bytes
+ *
+ * Description:
+ * Generate a CIPSO option using the restricted bitmap tag, tag type #1. The
+ * actual buffer length may be larger than the indicated size due to
+ * translation between host and network category bitmaps. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int cipso_v4_gentag_rbm(const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr,
+ unsigned char **buffer,
+ u32 *buffer_len)
+{
+ int ret_val = -EPERM;
+ unsigned char *buf = NULL;
+ u32 buf_len;
+ u32 level;
+
+ if (secattr->mls_cat) {
+ buf = kzalloc(CIPSO_V4_HDR_LEN + 4 + CIPSO_V4_TAG1_CAT_LEN,
+ GFP_ATOMIC);
+ if (buf == NULL)
+ return -ENOMEM;
+
+ ret_val = cipso_v4_map_cat_rbm_hton(doi_def,
+ secattr->mls_cat,
+ secattr->mls_cat_len,
+ &buf[CIPSO_V4_HDR_LEN + 4],
+ CIPSO_V4_TAG1_CAT_LEN);
+ if (ret_val < 0)
+ goto gentag_failure;
+
+ /* XXX - this will send packets using the "optimized" format
+ when possibile as specified in section 3.4.2.6 of the
+ CIPSO draft */
+ if (cipso_v4_rbm_optfmt && (ret_val > 0 && ret_val < 10))
+ ret_val = 10;
+
+ buf_len = 4 + ret_val;
+ } else {
+ buf = kzalloc(CIPSO_V4_HDR_LEN + 4, GFP_ATOMIC);
+ if (buf == NULL)
+ return -ENOMEM;
+ buf_len = 4;
+ }
+
+ ret_val = cipso_v4_map_lvl_hton(doi_def, secattr->mls_lvl, &level);
+ if (ret_val != 0)
+ goto gentag_failure;
+
+ ret_val = cipso_v4_gentag_hdr(doi_def, buf_len, buf);
+ if (ret_val != 0)
+ goto gentag_failure;
+
+ buf[CIPSO_V4_HDR_LEN] = 0x01;
+ buf[CIPSO_V4_HDR_LEN + 1] = buf_len;
+ buf[CIPSO_V4_HDR_LEN + 3] = level;
+
+ *buffer = buf;
+ *buffer_len = CIPSO_V4_HDR_LEN + buf_len;
+
+ return 0;
+
+gentag_failure:
+ kfree(buf);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_parsetag_rbm - Parse a CIPSO restricted bitmap tag
+ * @doi_def: the DOI definition
+ * @tag: the CIPSO tag
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Parse a CIPSO restricted bitmap tag (tag type #1) and return the security
+ * attributes in @secattr. Return zero on success, negatives values on
+ * failure.
+ *
+ */
+static int cipso_v4_parsetag_rbm(const struct cipso_v4_doi *doi_def,
+ const unsigned char *tag,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+ u8 tag_len = tag[1];
+ u32 level;
+
+ ret_val = cipso_v4_map_lvl_ntoh(doi_def, tag[3], &level);
+ if (ret_val != 0)
+ return ret_val;
+ secattr->mls_lvl = level;
+ secattr->mls_lvl_vld = 1;
+
+ if (tag_len > 4) {
+ switch (doi_def->type) {
+ case CIPSO_V4_MAP_PASS:
+ secattr->mls_cat_len = tag_len - 4;
+ break;
+ case CIPSO_V4_MAP_STD:
+ secattr->mls_cat_len =
+ doi_def->map.std->cat.local_size;
+ break;
+ }
+ secattr->mls_cat = kzalloc(secattr->mls_cat_len, GFP_ATOMIC);
+ if (secattr->mls_cat == NULL)
+ return -ENOMEM;
+
+ ret_val = cipso_v4_map_cat_rbm_ntoh(doi_def,
+ &tag[4],
+ tag_len - 4,
+ secattr->mls_cat,
+ secattr->mls_cat_len);
+ if (ret_val < 0) {
+ kfree(secattr->mls_cat);
+ return ret_val;
+ }
+ secattr->mls_cat_len = ret_val;
+ }
+
+ return 0;
+}
+
+/**
+ * cipso_v4_validate - Validate a CIPSO option
+ * @option: the start of the option, on error it is set to point to the error
+ *
+ * Description:
+ * This routine is called to validate a CIPSO option, it checks all of the
+ * fields to ensure that they are at least valid, see the draft snippet below
+ * for details. If the option is valid then a zero value is returned and
+ * the value of @option is unchanged. If the option is invalid then a
+ * non-zero value is returned and @option is adjusted to point to the
+ * offending portion of the option. From the IETF draft ...
+ *
+ * "If any field within the CIPSO options, such as the DOI identifier, is not
+ * recognized the IP datagram is discarded and an ICMP 'parameter problem'
+ * (type 12) is generated and returned. The ICMP code field is set to 'bad
+ * parameter' (code 0) and the pointer is set to the start of the CIPSO field
+ * that is unrecognized."
+ *
+ */
+int cipso_v4_validate(unsigned char **option)
+{
+ unsigned char *opt = *option;
+ unsigned char *tag;
+ unsigned char opt_iter;
+ unsigned char err_offset = 0;
+ unsigned char locked = 0;
+ u8 opt_len;
+ u8 tag_len;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 tag_iter;
+
+ /* XXX - caller already checks for length values that are too large */
+ opt_len = opt[1];
+ if (opt_len < 8) {
+ err_offset = 1;
+ goto validate_return;
+ }
+
+ rcu_read_lock();
+ locked = 1;
+ doi_def = cipso_v4_doi_getdef(ntohl(*((u32 *)&opt[2])));
+ if (doi_def == NULL) {
+ err_offset = 2;
+ goto validate_return;
+ }
+
+ opt_iter = 6;
+ tag = opt + opt_iter;
+ while (opt_iter < opt_len) {
+ for (tag_iter = 0; doi_def->tags[tag_iter] != tag[0];)
+ if (doi_def->tags[tag_iter] == CIPSO_V4_TAG_INVALID ||
+ ++tag_iter == CIPSO_V4_TAG_MAXCNT) {
+ err_offset = opt_iter;
+ goto validate_return;
+ }
+
+ tag_len = tag[1];
+ if (tag_len > (opt_len - opt_iter)) {
+ err_offset = opt_iter + 1;
+ goto validate_return;
+ }
+
+ switch (tag[0]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ if (tag_len < 4) {
+ err_offset = opt_iter + 1;
+ goto validate_return;
+ }
+
+ /* XXX - we are already going to do all the
+ verification necessary at the socket layer so from
+ our point of view it is safe to turn these checks
+ off (and less work), however, the CIPSO draft says
+ we should do all the CIPSO validations here but it
+ doesn't really specify _exactly_ what we need to
+ validate ... so, just make it a sysctl tunable */
+ if (cipso_v4_rbm_strictvalid) {
+ if (cipso_v4_map_lvl_valid(doi_def,
+ tag[3]) < 0) {
+ err_offset = opt_iter + 3;
+ goto validate_return;
+ }
+ if (tag_len > 4 &&
+ cipso_v4_map_cat_rbm_valid(doi_def,
+ &tag[4],
+ tag_len - 4) < 0) {
+ err_offset = opt_iter + 4;
+ goto validate_return;
+ }
+ }
+ break;
+ default:
+ err_offset = opt_iter;
+ goto validate_return;
+ }
+
+ tag += tag_len;
+ opt_iter += tag_len;
+ }
+
+validate_return:
+ if (locked)
+ rcu_read_unlock();
+ *option = opt + err_offset;
+ return err_offset;
+}
+
+/**
+ * cipso_v4_error - Send the correct reponse for a bad packet
+ * @skb: the packet
+ * @error: the error code
+ * @gateway: CIPSO gateway flag
+ *
+ * Description:
+ * Based on the error code given in @error, send an ICMP error message back to
+ * the originating host. From the IETF draft ...
+ *
+ * "If the contents of the CIPSO [option] are valid but the security label is
+ * outside of the configured host or port label range, the datagram is
+ * discarded and an ICMP 'destination unreachable' (type 3) is generated and
+ * returned. The code field of the ICMP is set to 'communication with
+ * destination network administratively prohibited' (code 9) or to
+ * 'communication with destination host administratively prohibited'
+ * (code 10). The value of the code is dependent on whether the originator
+ * of the ICMP message is acting as a CIPSO host or a CIPSO gateway. The
+ * recipient of the ICMP message MUST be able to handle either value. The
+ * same procedure is performed if a CIPSO [option] can not be added to an
+ * IP packet because it is too large to fit in the IP options area."
+ *
+ * "If the error is triggered by receipt of an ICMP message, the message is
+ * discarded and no response is permitted (consistent with general ICMP
+ * processing rules)."
+ *
+ */
+void cipso_v4_error(struct sk_buff *skb, const int error, const u32 gateway)
+{
+ if (skb->nh.iph->protocol == IPPROTO_ICMP || error != -EACCES)
+ return;
+
+ if (gateway)
+ icmp_send(skb, ICMP_DEST_UNREACH, ICMP_NET_ANO, 0);
+ else
+ icmp_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_ANO, 0);
+}
+
+/**
+ * cipso_v4_socket_setopt - Set the CIPSO option string on a socket
+ * @sock: the socket
+ * @opt: the option string
+ * @opt_len: the option string length
+ *
+ * Description:
+ * Attempt to add the given CIPSO option string to the socket. The caller is
+ * responsibile for locking and releasing the socket. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int cipso_v4_socket_setopt(struct socket *sock,
+ unsigned char *opt,
+ u32 opt_len)
+{
+ int ret_val;
+ struct ip_options *ip_opt = NULL;
+ struct sock *sk = sock->sk;
+ struct inet_sock *sk_inet;
+ struct inet_connection_sock *sk_conn;
+
+ ret_val = ip_options_get(&ip_opt, opt, opt_len);
+ if (ret_val != 0)
+ goto socket_setopt_return;
+ ip_opt->is_setbyuser = 0;
+ sk_inet = inet_sk(sk);
+ if (sk_inet->is_icsk) {
+ sk_conn = inet_csk(sk);
+ if (sk_inet->opt)
+ sk_conn->icsk_ext_hdr_len -= sk_inet->opt->optlen;
+ sk_conn->icsk_ext_hdr_len += ip_opt->optlen;
+ sk_conn->icsk_sync_mss(sk, sk_conn->icsk_pmtu_cookie);
+ }
+ ip_opt = xchg(&sk_inet->opt, ip_opt);
+
+ ret_val = 0;
+
+socket_setopt_return:
+ kfree(opt);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_socket_setattr - Add a CIPSO option to a socket
+ * @sock: the socket
+ * @doi_def: the CIPSO DOI to use
+ * @secattr: the specific security attributes of the socket
+ *
+ * Description:
+ * Set the CIPSO option on the given socket using the DOI definition and
+ * security attributes passed to the function. Returns zero on success and
+ * negative values on failure.
+ *
+ */
+int cipso_v4_socket_setattr(const struct socket *sock,
+ const struct cipso_v4_doi *doi_def,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -EPERM;
+ u32 iter;
+ unsigned char *buf = NULL;
+ u32 buf_len = 0;
+ struct ip_options *opt = NULL;
+ struct sock *sk;
+ struct inet_sock *sk_inet;
+ struct inet_connection_sock *sk_conn;
+
+ /* In the case of sock_create_lite(), the sock->sk field is not
+ defined yet but it is not a problem as the only users of these
+ "lite" PF_INET sockets are functions which do an accept() call
+ afterwards so we will label the socket as part of the accept(). */
+ sk = sock->sk;
+ if (sk == NULL)
+ return 0;
+
+ /* XXX - this code assumes only one tag per CIPSO option which isnt
+ really a good assumption to make but since we only support the MAC
+ tags right now it is a safe assumption */
+ iter = 0;
+ do {
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ ret_val = cipso_v4_gentag_rbm(doi_def,
+ secattr,
+ &buf,
+ &buf_len);
+ break;
+ default:
+ ret_val = -EPERM;
+ goto socket_setattr_failure;
+ }
+
+ iter++;
+ } while (ret_val != 0 &&
+ iter < CIPSO_V4_TAG_MAXCNT &&
+ doi_def->tags[iter] != CIPSO_V4_TAG_INVALID);
+ if (ret_val != 0)
+ goto socket_setattr_failure;
+
+ ret_val = ip_options_get(&opt, buf, buf_len);
+ if (ret_val != 0)
+ goto socket_setattr_failure;
+ opt->is_setbyuser = 0;
+
+ kfree(buf);
+ buf = NULL;
+
+ lock_sock(sk);
+ sk_inet = inet_sk(sk);
+ if (sk_inet->is_icsk) {
+ sk_conn = inet_csk(sk);
+ if (sk_inet->opt)
+ sk_conn->icsk_ext_hdr_len -= sk_inet->opt->optlen;
+ sk_conn->icsk_ext_hdr_len += opt->optlen;
+ sk_conn->icsk_sync_mss(sk, sk_conn->icsk_pmtu_cookie);
+ }
+ opt = xchg(&sk_inet->opt, opt);
+ release_sock(sk);
+ kfree(opt);
+
+ return 0;
+
+socket_setattr_failure:
+ kfree(buf);
+ kfree(opt);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_socket_getopt - Get the CIPSO IP option string from the socket
+ * @sock: the socket
+ * @opt: the option string
+ * @opt_len: the option string length in bytes
+ *
+ * Description:
+ * If a CIPSO option is present in the socket's IP options then return a
+ * a pointer to the option string. The caller is responsibile for locking
+ * and releasing the socket. Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int cipso_v4_socket_getopt(const struct socket *sock,
+ unsigned char **opt,
+ u32 *opt_len)
+{
+ struct ip_options *ip_opt = inet_sk(sock->sk)->opt;
+ unsigned char *cipso_ptr;
+
+ if (ip_opt == NULL || ip_opt->cipso == 0)
+ return -ENOMSG;
+ cipso_ptr = ip_opt->__data + ip_opt->cipso - sizeof(struct iphdr);
+
+ *opt = cipso_ptr;
+ *opt_len = cipso_ptr[1];
+
+ return 0;
+}
+
+/**
+ * cipso_v4_socket_getattr - Get the security attributes from a socket
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Query @sock to see if there is a CIPSO option attached to the socket and if
+ * there is return the CIPSO security attributes in @secattr. Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -ENOMSG;
+ struct sock *sk;
+ struct inet_sock *sk_inet;
+ unsigned char *cipso_ptr;
+ u32 doi;
+ struct cipso_v4_doi *doi_def;
+
+ sk = sock->sk;
+ lock_sock(sk);
+ sk_inet = inet_sk(sk);
+ if (sk_inet->opt == NULL || sk_inet->opt->cipso == 0)
+ goto socket_getattr_return;
+ cipso_ptr = sk_inet->opt->__data + sk_inet->opt->cipso -
+ sizeof(struct iphdr);
+ ret_val = cipso_v4_cache_check(cipso_ptr, cipso_ptr[1], secattr);
+ if (ret_val == 0)
+ goto socket_getattr_return;
+
+ doi = ntohl(*(u32 *)&cipso_ptr[2]);
+ rcu_read_lock();
+ doi_def = cipso_v4_doi_getdef(doi);
+ if (doi_def == NULL) {
+ rcu_read_unlock();
+ goto socket_getattr_return;
+ }
+ switch (cipso_ptr[6]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ ret_val = cipso_v4_parsetag_rbm(doi_def,
+ &cipso_ptr[6],
+ secattr);
+ break;
+ }
+ rcu_read_unlock();
+
+socket_getattr_return:
+ release_sock(sk);
+ return ret_val;
+}
+
+/**
+ * cipso_v4_skbuff_getattr - Get the security attributes from the CIPSO option
+ * @skb: the packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Parse the given packet's CIPSO option and return the security attributes.
+ * Returns zero on success and negative values on failure.
+ *
+ */
+int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -ENOMSG;
+ unsigned char *cipso_ptr;
+ u32 doi;
+ struct cipso_v4_doi *doi_def;
+
+ if (!CIPSO_V4_OPTEXIST(skb))
+ return -ENOMSG;
+ cipso_ptr = CIPSO_V4_OPTPTR(skb);
+ if (cipso_v4_cache_check(cipso_ptr, cipso_ptr[1], secattr) == 0)
+ return 0;
+
+ doi = ntohl(*(u32 *)&cipso_ptr[2]);
+ rcu_read_lock();
+ doi_def = cipso_v4_doi_getdef(doi);
+ if (doi_def == NULL)
+ goto skbuff_getattr_return;
+ switch (cipso_ptr[6]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ ret_val = cipso_v4_parsetag_rbm(doi_def,
+ &cipso_ptr[6],
+ secattr);
+ break;
+ }
+
+skbuff_getattr_return:
+ rcu_read_unlock();
+ return ret_val;
+}
+
+/*
+ * Setup Functions
+ */
+
+/**
+ * cipso_v4_init - Initialize the CIPSO module
+ *
+ * Description:
+ * Initialize the CIPSO module and prepare it for use. Returns zero on success
+ * and negative values on failure.
+ *
+ */
+static int __init cipso_v4_init(void)
+{
+ return cipso_v4_cache_init(CIPSO_V4_CACHE_BUCKETS);
+}
+
+/**
+ * cipso_v4_exit - Cleanup after the CIPSO module
+ *
+ * Description:
+ * Perform any necessary cleanup tasks before the CIPSO module goes away.
+ *
+ */
+static void __exit cipso_v4_exit(void)
+{
+ cipso_v4_cache_destroy();
+}
+
+subsys_initcall(cipso_v4_init);
Index: linux-2.6.18-rc1/net/ipv4/sysctl_net_ipv4.c
===================================================================
--- linux-2.6.18-rc1.orig/net/ipv4/sysctl_net_ipv4.c
+++ linux-2.6.18-rc1/net/ipv4/sysctl_net_ipv4.c
@@ -17,6 +17,7 @@
#include <net/ip.h>
#include <net/route.h>
#include <net/tcp.h>
+#include <net/cipso_ipv4.h>
/* From af_inet.c */
extern int sysctl_ip_nonlocal_bind;
@@ -697,6 +698,40 @@ ctl_table ipv4_table[] = {
.mode = 0644,
.proc_handler = &proc_dointvec
},
+#ifdef CONFIG_NETLABEL
+ {
+ .ctl_name = NET_CIPSOV4_CACHE_ENABLE,
+ .procname = "cipso_cache_enable",
+ .data = &cipso_v4_cache_enabled,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
+ .ctl_name = NET_CIPSOV4_CACHE_BUCKET_SIZE,
+ .procname = "cipso_cache_bucket_size",
+ .data = &cipso_v4_cache_bucketsize,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
+ .ctl_name = NET_CIPSOV4_RBM_OPTFMT,
+ .procname = "cipso_rbm_optfmt",
+ .data = &cipso_v4_rbm_optfmt,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
+ .ctl_name = NET_CIPSOV4_RBM_STRICTVALID,
+ .procname = "cipso_rbm_strictvalid",
+ .data = &cipso_v4_rbm_strictvalid,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+#endif /* CONFIG_NETLABEL */
{ .ctl_name = 0 }
};
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 4/7] NetLabel: core NetLabel subsystem
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-core --]
[-- Type: text/plain, Size: 78332 bytes --]
Add a new kernel subsystem, NetLabel, to provide explicit packet labeling
services (CIPSO, RIPSO, etc.) to LSM developers. NetLabel is designed to work
in conjunction with a LSM to intercept and decode security labels on incoming
network packets as well as ensure that outgoing network packets are labeled
according to the security mechanism employed by the LSM. The NetLabel
subsystem is configured through a Generic NETLINK interface described in the
header files included in this patch.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/Makefile | 1
net/netlabel/Kconfig | 14
net/netlabel/Makefile | 16
net/netlabel/netlabel_cipso_v4.h | 218 ++++++++++++
net/netlabel/netlabel_domainhash.c | 560 +++++++++++++++++++++++++++++++
net/netlabel/netlabel_domainhash.h | 64 +++
net/netlabel/netlabel_kapi.c | 299 ++++++++++++++++
net/netlabel/netlabel_mgmt.c | 661 +++++++++++++++++++++++++++++++++++++
net/netlabel/netlabel_mgmt.h | 250 +++++++++++++
net/netlabel/netlabel_unlabeled.h | 99 +++++
net/netlabel/netlabel_user.c | 181 ++++++++++
net/netlabel/netlabel_user.h | 210 +++++++++++
12 files changed, 2573 insertions(+)
Index: linux-2.6.18-rc1/net/Makefile
===================================================================
--- linux-2.6.18-rc1.orig/net/Makefile
+++ linux-2.6.18-rc1/net/Makefile
@@ -46,6 +46,7 @@ obj-$(CONFIG_IP_DCCP) += dccp/
obj-$(CONFIG_IP_SCTP) += sctp/
obj-$(CONFIG_IEEE80211) += ieee80211/
obj-$(CONFIG_TIPC) += tipc/
+obj-$(CONFIG_NETLABEL) += netlabel/
ifeq ($(CONFIG_NET),y)
obj-$(CONFIG_SYSCTL) += sysctl_net.o
Index: linux-2.6.18-rc1/net/netlabel/Kconfig
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/Kconfig
@@ -0,0 +1,14 @@
+#
+# NetLabel configuration
+#
+
+config NETLABEL
+ bool "NetLabel subsystem support"
+ depends on NET && SECURITY
+ default n
+ ---help---
+ NetLabel provides support for explicit network packet labeling
+ protocols such as CIPSO and RIPSO. For more information see
+ Documentation/netlabel.
+
+ If you are unsure, say N.
Index: linux-2.6.18-rc1/net/netlabel/Makefile
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/Makefile
@@ -0,0 +1,16 @@
+#
+# Makefile for the NetLabel subsystem.
+#
+# Feb 9, 2006, Paul Moore <paul.moore@hp.com>
+#
+
+# base objects
+obj-y := netlabel_user.o netlabel_kapi.o netlabel_domainhash.o
+
+# management objects
+obj-y += netlabel_mgmt.o
+
+# protocol modules
+obj-y += netlabel_unlabeled.o
+obj-y += netlabel_cipso_v4.o
+
Index: linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.h
@@ -0,0 +1,218 @@
+/*
+ * NetLabel CIPSO/IPv4 Support
+ *
+ * This file defines the CIPSO/IPv4 functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_CIPSO_V4
+#define _NETLABEL_CIPSO_V4
+
+#include <net/netlabel.h>
+
+/*
+ * The following NetLabel payloads are supported by the CIPSO subsystem, all
+ * of which are preceeded by the nlmsghdr struct.
+ *
+ * o ACK:
+ * Sent by the kernel in response to an applications message, applications
+ * should never send this message.
+ *
+ * +----------------------+-----------------------+
+ * | seq number (32 bits) | return code (32 bits) |
+ * +----------------------+-----------------------+
+ *
+ * seq number: the sequence number of the original message, taken from the
+ * nlmsghdr structure
+ * return code: return value, based on errno values
+ *
+ * o ADD:
+ * Sent by an application to add a new DOI mapping table, after completion
+ * of the task the kernel should ACK this message.
+ *
+ * +---------------+--------------------+---------------------+
+ * | DOI (32 bits) | map type (32 bits) | tag count (32 bits) | ...
+ * +---------------+--------------------+---------------------+
+ *
+ * +-----------------+
+ * | tag #X (8 bits) | ... repeated
+ * +-----------------+
+ *
+ * +-------------- ---- --- -- -
+ * | mapping data
+ * +-------------- ---- --- -- -
+ *
+ * DOI: the DOI value
+ * map type: the mapping table type (defined in the cipso_ipv4.h header
+ * as CIPSO_V4_MAP_*)
+ * tag count: the number of tags, must be greater than zero
+ * tag: the CIPSO tag for the DOI, tags listed first are given
+ * higher priorirty when sending packets
+ * mapping data: specific to the map type (see below)
+ *
+ * CIPSO_V4_MAP_STD
+ *
+ * +------------------+-----------------------+----------------------+
+ * | levels (32 bits) | max l level (32 bits) | max r level (8 bits) | ...
+ * +------------------+-----------------------+----------------------+
+ *
+ * +----------------------+---------------------+---------------------+
+ * | categories (32 bits) | max l cat (32 bits) | max r cat (16 bits) | ...
+ * +----------------------+---------------------+---------------------+
+ *
+ * +--------------------------+-------------------------+
+ * | local level #X (32 bits) | CIPSO level #X (8 bits) | ... repeated
+ * +--------------------------+-------------------------+
+ *
+ * +-----------------------------+-----------------------------+
+ * | local category #X (32 bits) | CIPSO category #X (16 bits) | ... repeated
+ * +-----------------------------+-----------------------------+
+ *
+ * levels: the number of level mappings
+ * max l level: the highest local level
+ * max r level: the highest remote/CIPSO level
+ * categories: the number of category mappings
+ * max l cat: the highest local category
+ * max r cat: the highest remote/CIPSO category
+ * local level: the local part of a level mapping
+ * CIPSO level: the remote/CIPSO part of a level mapping
+ * local category: the local part of a category mapping
+ * CIPSO category: the remote/CIPSO part of a category mapping
+ *
+ * CIPSO_V4_MAP_PASS
+ *
+ * No mapping data is needed for this map type.
+ *
+ * o REMOVE:
+ * Sent by an application to remove a specific DOI mapping table from the
+ * CIPSO V4 system. The kernel should ACK this message.
+ *
+ * +---------------+
+ * | DOI (32 bits) |
+ * +---------------+
+ *
+ * DOI: the DOI value
+ *
+ * o LIST:
+ * Sent by an application to list the details of a DOI definition. The
+ * kernel should send an ACK on error or a response as indicated below. The
+ * application generated message format is shown below.
+ *
+ * +---------------+
+ * | DOI (32 bits) |
+ * +---------------+
+ *
+ * DOI: the DOI value
+ *
+ * The valid response message format depends on the type of the DOI mapping,
+ * the known formats are shown below.
+ *
+ * +--------------------+
+ * | map type (32 bits) | ...
+ * +--------------------+
+ *
+ * map type: the DOI mapping table type (defined in the cipso_ipv4.h
+ * header as CIPSO_V4_MAP_*)
+ *
+ * (map type == CIPSO_V4_MAP_STD)
+ *
+ * +----------------+------------------+----------------------+
+ * | tags (32 bits) | levels (32 bits) | categories (32 bits) | ...
+ * +----------------+------------------+----------------------+
+ *
+ * +-----------------+
+ * | tag #X (8 bits) | ... repeated
+ * +-----------------+
+ *
+ * +--------------------------+-------------------------+
+ * | local level #X (32 bits) | CIPSO level #X (8 bits) | ... repeated
+ * +--------------------------+-------------------------+
+ *
+ * +-----------------------------+-----------------------------+
+ * | local category #X (32 bits) | CIPSO category #X (16 bits) | ... repeated
+ * +-----------------------------+-----------------------------+
+ *
+ * tags: the number of CIPSO tag types
+ * levels: the number of level mappings
+ * categories: the number of category mappings
+ * tag: the tag number, tags listed first are given higher
+ * priority when sending packets
+ * local level: the local part of a level mapping
+ * CIPSO level: the remote/CIPSO part of a level mapping
+ * local category: the local part of a category mapping
+ * CIPSO category: the remote/CIPSO part of a category mapping
+ *
+ * (map type == CIPSO_V4_MAP_PASS)
+ *
+ * +----------------+
+ * | tags (32 bits) | ...
+ * +----------------+
+ *
+ * +-----------------+
+ * | tag #X (8 bits) | ... repeated
+ * +-----------------+
+ *
+ * tags: the number of CIPSO tag types
+ * tag: the tag number, tags listed first are given higher
+ * priority when sending packets
+ *
+ * o LISTALL:
+ * This message is sent by an application to list the valid DOIs on the
+ * system. There is no payload and the kernel should respond with an ACK
+ * or the following message.
+ *
+ * +---------------------+------------------+-----------------------+
+ * | DOI count (32 bits) | DOI #X (32 bits) | map type #X (32 bits) |
+ * +---------------------+------------------+-----------------------+
+ *
+ * +-----------------------+
+ * | map type #X (32 bits) | ...
+ * +-----------------------+
+ *
+ * DOI count: the number of DOIs
+ * DOI: the DOI value
+ * map type: the DOI mapping table type (defined in the cipso_ipv4.h
+ * header as CIPSO_V4_MAP_*)
+ *
+ */
+
+/* NetLabel CIPSOv4 commands */
+enum {
+ NLBL_CIPSOV4_C_UNSPEC,
+ NLBL_CIPSOV4_C_ACK,
+ NLBL_CIPSOV4_C_ADD,
+ NLBL_CIPSOV4_C_REMOVE,
+ NLBL_CIPSOV4_C_LIST,
+ NLBL_CIPSOV4_C_LISTALL,
+ __NLBL_CIPSOV4_C_MAX,
+};
+#define NLBL_CIPSOV4_C_MAX (__NLBL_CIPSOV4_C_MAX - 1)
+
+/* NetLabel protocol functions */
+int netlbl_cipsov4_genl_init(void);
+void netlbl_cipsov4_genl_exit(void);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.c
@@ -0,0 +1,560 @@
+/*
+ * NetLabel Domain Hash Table
+ *
+ * This file manages the domain hash table that NetLabel uses to determine
+ * which network labeling protocol to use for a given domain. The NetLabel
+ * system manages static and dynamic label mappings for network protocols such
+ * as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/skbuff.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+#include "netlabel_mgmt.h"
+#include "netlabel_domainhash.h"
+
+struct netlbl_domhsh_tbl {
+ struct list_head *tbl;
+ u32 size;
+};
+
+/* Domain hash table */
+/* XXX - updates should be so rare that having one spinlock for the entire
+ hash table should be okay */
+DEFINE_SPINLOCK(netlbl_domhsh_lock);
+static struct netlbl_domhsh_tbl *netlbl_domhsh = NULL;
+
+/* Default domain mapping */
+DEFINE_SPINLOCK(netlbl_domhsh_def_lock);
+static struct netlbl_dom_map *netlbl_domhsh_def = NULL;
+
+/*
+ * Domain Hash Table Helper Functions
+ */
+
+/**
+ * netlbl_domhsh_free_entry - Frees a domain hash table entry
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to a hash table entry can be released
+ * safely.
+ *
+ */
+static void netlbl_domhsh_free_entry(struct rcu_head *entry)
+{
+ struct netlbl_dom_map *ptr;
+
+ ptr = container_of(entry, struct netlbl_dom_map, rcu);
+ kfree(ptr->domain);
+ kfree(ptr);
+}
+
+/**
+ * netlbl_domhsh_hash - Hashing function for the domain hash table
+ * @domain: the domain name to hash
+ *
+ * Description:
+ * This is the hashing function for the domain hash table, it returns the
+ * correct bucket number for the domain. The caller is responsibile for
+ * calling the rcu_read_[un]lock() functions.
+ *
+ */
+static u32 netlbl_domhsh_hash(const char *key)
+{
+ u32 iter;
+ u32 val;
+ u32 len;
+
+ /* This is taken (with slight modification) from
+ security/selinux/ss/symtab.c:symhash() */
+
+ for (iter = 0, val = 0, len = strlen(key); iter < len; iter++)
+ val = (val << 4 | (val >> (8 * sizeof(u32) - 4))) ^ key[iter];
+ return val & (rcu_dereference(netlbl_domhsh)->size - 1);
+}
+
+/**
+ * netlbl_domhsh_search - Search for a domain entry
+ * @domain: the domain
+ * @def: return default if no match is found
+ *
+ * Description:
+ * Searches the domain hash table and returns a pointer to the hash table
+ * entry if found, otherwise NULL is returned. If @def is non-zero and a
+ * match is not found in the domain hash table the default mapping is returned
+ * if it exists. The caller is responsibile for the rcu hash table locks
+ * (i.e. the caller much call rcu_read_[un]lock()).
+ *
+ */
+static struct netlbl_dom_map *netlbl_domhsh_search(const char *domain,
+ const u32 def)
+{
+ u32 bkt;
+ struct netlbl_dom_map *iter;
+
+ if (domain != NULL) {
+ bkt = netlbl_domhsh_hash(domain);
+ list_for_each_entry_rcu(iter, &netlbl_domhsh->tbl[bkt], list)
+ if (iter->valid && strcmp(iter->domain, domain) == 0)
+ return iter;
+ }
+
+ if (def != 0) {
+ iter = rcu_dereference(netlbl_domhsh_def);
+ if (iter != NULL && iter->valid)
+ return iter;
+ }
+
+ return NULL;
+}
+
+/*
+ * Domain Hash Table Functions
+ */
+
+/**
+ * netlbl_domhsh_init - Init for the domain hash
+ * @size: the number of bits to use for the hash buckets
+ *
+ * Description:
+ * Initializes the domain hash table, should be called only by
+ * netlbl_user_init() during module load. Returns zero on success, non-zero
+ * values on error.
+ *
+ */
+int netlbl_domhsh_init(const u32 size)
+{
+ u32 iter;
+ struct netlbl_domhsh_tbl *hsh_tbl;
+
+ if (size == 0)
+ return -EINVAL;
+
+ hsh_tbl = kmalloc(sizeof(*hsh_tbl), GFP_KERNEL);
+ if (hsh_tbl == NULL)
+ return -ENOMEM;
+ hsh_tbl->size = 1 << size;
+ hsh_tbl->tbl = kcalloc(hsh_tbl->size,
+ sizeof(struct list_head),
+ GFP_KERNEL);
+ if (hsh_tbl->tbl == NULL) {
+ kfree(hsh_tbl);
+ return -ENOMEM;
+ }
+ for (iter = 0; iter < hsh_tbl->size; iter++)
+ INIT_LIST_HEAD(&hsh_tbl->tbl[iter]);
+
+ rcu_read_lock();
+ spin_lock(&netlbl_domhsh_lock);
+ rcu_assign_pointer(netlbl_domhsh, hsh_tbl);
+ spin_unlock(&netlbl_domhsh_lock);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * netlbl_domhsh_exit - Cleanup for the domain hash
+ *
+ * Description:
+ * Clears and frees the domain hash table. No real serious effort is made
+ * towards locking but that shouldn't be an issue since we should only be
+ * called when the module is being unloaded. Returns zero on success, non-zero
+ * values on error.
+ *
+ */
+void netlbl_domhsh_exit(void)
+{
+ u32 iter_bkt;
+ struct netlbl_domhsh_tbl *hsh_tbl;
+ struct netlbl_dom_map *iter_list;
+
+ rcu_read_lock();
+ if (rcu_dereference(netlbl_domhsh) == NULL) {
+ rcu_read_unlock();
+ return;
+ }
+ iter_list = rcu_dereference(netlbl_domhsh_def);
+ spin_lock(&netlbl_domhsh_def_lock);
+ rcu_assign_pointer(netlbl_domhsh_def, NULL);
+ spin_unlock(&netlbl_domhsh_def_lock);
+ rcu_read_unlock();
+
+ if (iter_list) {
+ synchronize_rcu();
+ netlbl_domhsh_free_entry(&iter_list->rcu);
+ }
+
+ rcu_read_lock();
+ hsh_tbl = rcu_dereference(netlbl_domhsh);
+ spin_lock(&netlbl_domhsh_lock);
+ rcu_assign_pointer(netlbl_domhsh, NULL);
+ spin_unlock(&netlbl_domhsh_lock);
+ rcu_read_unlock();
+
+ synchronize_rcu();
+
+ /* don't use the rcu functions here since the table is no longer in use
+ outside this function's scope */
+ for (iter_bkt = 0; iter_bkt < hsh_tbl->size; iter_bkt++)
+ list_for_each_entry(iter_list, &hsh_tbl->tbl[iter_bkt], list) {
+ iter_list->valid = 0;
+ list_del(&iter_list->list);
+ netlbl_domhsh_free_entry(&iter_list->rcu);
+ }
+ kfree(hsh_tbl->tbl);
+ kfree(hsh_tbl);
+}
+
+/**
+ * netlbl_domhsh_add - Adds a entry to the domain hash table
+ * @entry: the entry to add
+ *
+ * Description:
+ * Adds a new entry to the domain hash table and handles any updates to the
+ * lower level protocol handler (i.e. CIPSO). Returns zero on success,
+ * negative on failure.
+ *
+ */
+int netlbl_domhsh_add(struct netlbl_dom_map *entry)
+{
+ int ret_val = -EPERM;
+ u32 bkt;
+
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ ret_val = 0;
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ ret_val = cipso_v4_doi_domhsh_add(entry->type_def.cipsov4,
+ entry->domain);
+ break;
+ }
+ if (ret_val != 0)
+ return ret_val;
+
+ entry->valid = 1;
+ INIT_RCU_HEAD(&entry->rcu);
+
+ rcu_read_lock();
+ if (entry->domain != NULL &&
+ netlbl_domhsh_search(entry->domain, 0) == NULL) {
+ bkt = netlbl_domhsh_hash(entry->domain);
+ spin_lock(&netlbl_domhsh_lock);
+ list_add_tail_rcu(&entry->list, &netlbl_domhsh->tbl[bkt]);
+ spin_unlock(&netlbl_domhsh_lock);
+ } else if (entry->domain == NULL &&
+ rcu_dereference(netlbl_domhsh_def) == NULL) {
+ INIT_LIST_HEAD(&entry->list);
+ spin_lock(&netlbl_domhsh_def_lock);
+ rcu_assign_pointer(netlbl_domhsh_def, entry);
+ spin_unlock(&netlbl_domhsh_def_lock);
+ } else {
+ rcu_read_unlock();
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ BUG_ON(cipso_v4_doi_domhsh_remove(
+ entry->type_def.cipsov4,
+ entry->domain));
+ break;
+ }
+ return -EPERM;
+ }
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * netlbl_domhsh_add_default - Adds the default entry to the domain hash table
+ * @entry: the entry to add
+ *
+ * Description:
+ * Adds a new default entry to the domain hash table and handles any updates
+ * to the lower level protocol handler (i.e. CIPSO). Returns zero on success,
+ * negative on failure.
+ *
+ */
+int netlbl_domhsh_add_default(struct netlbl_dom_map *entry)
+{
+ return netlbl_domhsh_add(entry);
+}
+
+/**
+ * netlbl_domhsh_remove - Removes an entry from the domain hash table
+ * @domain: the domain to remove
+ *
+ * Description:
+ * Removes an entry from the domain hash table and handles any updates to the
+ * lower level protocol handler (i.e. CIPSO). Returns zero on success,
+ * negative on failure.
+ *
+ */
+int netlbl_domhsh_remove(const char *domain)
+{
+ int ret_val = -EPERM;
+ struct netlbl_dom_map *entry;
+
+ rcu_read_lock();
+ if (domain != NULL)
+ entry = netlbl_domhsh_search(domain, 0);
+ else
+ entry = netlbl_domhsh_search(domain, 1);
+ if (entry == NULL)
+ goto remove_return;
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (cipso_v4_doi_domhsh_remove(entry->type_def.cipsov4,
+ entry->domain) != 0)
+ goto remove_return;
+ break;
+ }
+ if (entry != rcu_dereference(netlbl_domhsh_def)) {
+ spin_lock(&netlbl_domhsh_lock);
+ entry->valid = 0;
+ list_del_rcu(&entry->list);
+ spin_unlock(&netlbl_domhsh_lock);
+ } else {
+ spin_lock(&netlbl_domhsh_def_lock);
+ entry->valid = 0;
+ rcu_assign_pointer(netlbl_domhsh_def, NULL);
+ spin_unlock(&netlbl_domhsh_def_lock);
+ }
+ call_rcu(&entry->rcu, netlbl_domhsh_free_entry);
+
+ ret_val = 0;
+
+remove_return:
+ rcu_read_unlock();
+ return ret_val;
+}
+
+/**
+ * netlbl_domhsh_remove_default - Removes the default entry from the table
+ *
+ * Description:
+ * Removes/resets the default entry for the domain hash table and handles any
+ * updates to the lower level protocol handler (i.e. CIPSO). Returns zero on
+ * success, non-zero on failure.
+ *
+ */
+int netlbl_domhsh_remove_default(void)
+{
+ return netlbl_domhsh_remove(NULL);
+}
+
+/**
+ * netlbl_domhsh_getentry - Get an entry from the domain hash table
+ * @domain: the domain name to search for
+ *
+ * Description:
+ * Look through the domain hash table searching for an entry to match @domain,
+ * return a pointer to a copy of the entry or NULL. The caller is responsibile
+ * for ensuring that rcu_read_[un]lock() is called.
+ *
+ */
+struct netlbl_dom_map *netlbl_domhsh_getentry(const char *domain)
+{
+ return netlbl_domhsh_search(domain, 1);
+}
+
+/**
+ * netlbl_domhsh_dump - Dump the domain hash table into a sk_buff
+ *
+ * Description:
+ * Dump the domain hash table into a buffer suitable for returning to an
+ * application in response to a NetLabel management DOMAIN message. This
+ * function may fail if another process is growing the hash table at the same
+ * time. The returned sk_buff has room at the front of the sk_buff for
+ * @headroom bytes. See netlabel.h for the DOMAIN message format. Returns a
+ * pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *netlbl_domhsh_dump(const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ ssize_t buf_len;
+ u32 bkt_iter;
+ u32 dom_cnt = 0;
+ struct netlbl_domhsh_tbl *hsh_tbl;
+ struct netlbl_dom_map *list_iter;
+ ssize_t tmp_len;
+
+ /* XXX - This is kinda ugly as we have to go through the table once to
+ determine how large of a buffer we need, drop the locks, allocate
+ the buffer, grab the locks, and finally fill the buffer. The
+ problem is that there is that open window where the table could
+ grow and we will end up short on space. */
+
+ buf_len = NETLBL_LEN_U32;
+ rcu_read_lock();
+ hsh_tbl = rcu_dereference(netlbl_domhsh);
+ for (bkt_iter = 0; bkt_iter < hsh_tbl->size; bkt_iter++)
+ list_for_each_entry_rcu(list_iter,
+ &hsh_tbl->tbl[bkt_iter], list) {
+ buf_len += 2 * NETLBL_LEN_U32 +
+ netlbl_align(strlen(list_iter->domain) + 1);
+ switch (list_iter->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ buf_len += 2 * NETLBL_LEN_U32;
+ break;
+ }
+ dom_cnt++;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+ netlbl_putinc_u32(&buf, dom_cnt, &buf_len);
+
+ rcu_read_lock();
+ hsh_tbl = rcu_dereference(netlbl_domhsh);
+ for (bkt_iter = 0; bkt_iter < hsh_tbl->size; bkt_iter++)
+ list_for_each_entry_rcu(list_iter,
+ &hsh_tbl->tbl[bkt_iter], list) {
+ tmp_len = netlbl_align(strlen(list_iter->domain) + 1);
+ if (buf_len < 2 * NETLBL_LEN_U32 + tmp_len)
+ goto dump_failure;
+ netlbl_putinc_u32(&buf, tmp_len, &buf_len);
+ netlbl_putinc_str(&buf, list_iter->domain, &buf_len);
+ netlbl_putinc_u32(&buf, list_iter->type, &buf_len);
+ switch (list_iter->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto dump_failure;
+ netlbl_putinc_u32(
+ &buf,
+ list_iter->type_def.cipsov4->type,
+ &buf_len);
+ netlbl_putinc_u32(
+ &buf,
+ list_iter->type_def.cipsov4->doi,
+ &buf_len);
+ break;
+ }
+ }
+ rcu_read_unlock();
+
+ return skb;
+
+dump_failure:
+ rcu_read_unlock();
+ kfree_skb(skb);
+ return NULL;
+}
+
+/**
+ * netlbl_domhsh_dump_default - Dump the default domain mapping into a sk_buff
+ *
+ * Description:
+ * Dump the default domain mapping into a buffer suitable for returning to an
+ * application in response to a NetLabel management DEFDOMAIN message. This
+ * function may fail if another process is changing the default domain mapping
+ * at the same time. The returned sk_buff has room at the front of the
+ * skb_buff for @headroom bytes. See netlabel.h for the DEFDOMAIN message
+ * format. Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *netlbl_domhsh_dump_default(const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ ssize_t buf_len;
+ struct netlbl_dom_map *entry;
+
+ /* XXX - This is kinda ugly as we have to go look at the default
+ mapping once to determine how large of a buffer we need, drop the
+ locks, allocate the buffer, grab the locks, and finally fill the
+ buffer. The problem is that there is that open window where the
+ default mapping could change on us, if that is the case we fail. */
+
+ buf_len = NETLBL_LEN_U32;
+ rcu_read_lock();
+ entry = rcu_dereference(netlbl_domhsh_def);
+ if (entry != NULL)
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ buf_len += 2 * NETLBL_LEN_U32;
+ break;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+
+ rcu_read_lock();
+ if (entry != rcu_dereference(netlbl_domhsh_def))
+ goto dump_default_failure;
+ if (entry != NULL) {
+ netlbl_putinc_u32(&buf, entry->type, &buf_len);
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto dump_default_failure;
+ netlbl_putinc_u32(&buf,
+ entry->type_def.cipsov4->type,
+ &buf_len);
+ netlbl_putinc_u32(&buf,
+ entry->type_def.cipsov4->doi,
+ &buf_len);
+ break;
+ }
+ } else
+ netlbl_putinc_u32(&buf, NETLBL_NLTYPE_NONE, &buf_len);
+ rcu_read_unlock();
+
+ return skb;
+
+dump_default_failure:
+ rcu_read_unlock();
+ kfree_skb(skb);
+ return NULL;
+}
Index: linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.h
@@ -0,0 +1,64 @@
+/*
+ * NetLabel Domain Hash Table
+ *
+ * This file manages the domain hash table that NetLabel uses to determine
+ * which network labeling protocol to use for a given domain. The NetLabel
+ * system manages static and dynamic label mappings for network protocols such
+ * as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_DOMAINHASH_H
+#define _NETLABEL_DOMAINHASH_H
+
+/* Domain hash table size */
+/* XXX - currently this number is an uneducated guess */
+#define NETLBL_DOMHSH_BITSIZE 7
+
+/* Domain mapping definition struct */
+struct netlbl_dom_map {
+ char *domain;
+ u32 type;
+ union {
+ struct cipso_v4_doi *cipsov4;
+ } type_def;
+
+ u32 valid;
+ struct list_head list;
+ struct rcu_head rcu;
+};
+
+/* init/exit functions */
+int netlbl_domhsh_init(const u32 size);
+void netlbl_domhsh_exit(void);
+
+/* Manipulate the domain hash table */
+int netlbl_domhsh_add(struct netlbl_dom_map *entry);
+int netlbl_domhsh_add_default(struct netlbl_dom_map *entry);
+int netlbl_domhsh_remove_default(void);
+struct netlbl_dom_map *netlbl_domhsh_getentry(const char *domain);
+struct sk_buff *netlbl_domhsh_dump(const size_t headroom);
+struct sk_buff *netlbl_domhsh_dump_default(const size_t headroom);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_kapi.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_kapi.c
@@ -0,0 +1,299 @@
+/*
+ * NetLabel Kernel API
+ *
+ * This file defines the kernel API for the NetLabel system. The NetLabel
+ * system manages static and dynamic label mappings for network protocols such
+ * as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <net/ip.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+#include "netlabel_domainhash.h"
+#include "netlabel_unlabeled.h"
+#include "netlabel_user.h"
+
+/*
+ * LSM Functions
+ */
+
+/**
+ * netlbl_socket_setattr - Label a socket using the correct protocol
+ * @sock: the socket to label
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Attach the correct label to the given socket using the security attributes
+ * specified in @secattr. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_socket_setattr(const struct socket *sock,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -ENOENT;
+ struct netlbl_dom_map *dom_entry;
+
+ rcu_read_lock();
+ dom_entry = netlbl_domhsh_getentry(secattr->domain);
+ if (dom_entry == NULL)
+ goto socket_setattr_return;
+ switch (dom_entry->type) {
+ case NETLBL_NLTYPE_CIPSOV4:
+ ret_val = cipso_v4_socket_setattr(sock,
+ dom_entry->type_def.cipsov4,
+ secattr);
+ break;
+ case NETLBL_NLTYPE_UNLABELED:
+ ret_val = 0;
+ break;
+ default:
+ ret_val = -ENOENT;
+ }
+
+socket_setattr_return:
+ rcu_read_unlock();
+ return ret_val;
+}
+
+/**
+ * netlbl_socket_peekattr - Get the security attributes on a queued packet
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Peek at the incoming packet queue of @sock and return the security
+ * attributes of the first packet in the queue. If there are no packets in
+ * the queue or no packets with security attributes return -ENOMSG. Otherwise
+ * return zero on success, negative values on failure.
+ *
+ */
+int netlbl_socket_peekattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+ struct sk_buff_head *sock_queue;
+ struct sk_buff *skb;
+ struct sk_buff *skb_tmp = NULL;
+
+ sock_queue = &sock->sk->sk_receive_queue;
+ /* XXX - We could just take the lock here and do a skb_peek() but we
+ might end up holding the lock for a long time so grab the
+ lock, do a peek, and clone the result */
+ spin_lock(&sock_queue->lock);
+ skb = skb_peek(sock_queue);
+ if (skb != NULL)
+ skb_tmp = skb_clone(skb, GFP_ATOMIC);
+ spin_unlock(&sock_queue->lock);
+ if (skb_tmp != NULL) {
+ ret_val = netlbl_skbuff_getattr(skb_tmp, secattr);
+ kfree_skb(skb_tmp);
+ } else
+ ret_val = -ENOMEM;
+
+ return ret_val;
+}
+
+/**
+ * netlbl_socket_getattr - Determine the security attributes of a socket
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Examines the given socket to see any NetLabel style labeling has been
+ * applied to the socket, if so it parses the socket label and returns the
+ * security attributes in @secattr. Returns zero on success, negative values
+ * on failure.
+ *
+ */
+int netlbl_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+
+ ret_val = cipso_v4_socket_getattr(sock, secattr);
+ if (ret_val == 0)
+ return 0;
+
+ return netlbl_unlabel_getattr(secattr);
+}
+
+/**
+ * netlbl_skbuff_getattr - Determine the security attributes of a packet
+ * @skb: the packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Examines the given packet to see if a recognized form of packet labeling
+ * is present, if so it parses the packet label and returns the security
+ * attributes in @secattr. Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int netlbl_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+
+ ret_val = cipso_v4_skbuff_getattr(skb, secattr);
+ if (ret_val == 0)
+ return 0;
+
+ return netlbl_unlabel_getattr(secattr);
+}
+
+/**
+ * netlbl_skbuff_err - Handle a LSM error on a sk_buff
+ * @skb: the packet
+ * @error: the error code
+ *
+ * Description:
+ * Deal with a LSM problem when handling the packet in @skb, typically this is
+ * a permission denied problem (-EACCES). The correct action is determined
+ * according to the packet's labeling protocol.
+ *
+ */
+void netlbl_skbuff_err(struct sk_buff *skb, int error)
+{
+ if (CIPSO_V4_OPTEXIST(skb))
+ cipso_v4_error(skb, error, 0);
+}
+
+/**
+ * netlbl_cache_invalidate - Invalidate all of the NetLabel protocol caches
+ *
+ * Description:
+ * For all of the NetLabel protocols that support some form of label mapping
+ * cache, invalidate the cache. Returns zero on success, negative values on
+ * error.
+ *
+ */
+void netlbl_cache_invalidate(void)
+{
+ cipso_v4_cache_invalidate();
+}
+
+/**
+ * netlbl_cache_add - Add an entry to a NetLabel protocol cache
+ * @skb: the packet
+ * @secattr: the packet's security attributes
+ *
+ * Description:
+ * Add the LSM security attributes for the given packet to the underlying
+ * NetLabel protocol's label mapping cache. Returns zero on success, negative
+ * values on error.
+ *
+ */
+int netlbl_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ if (secattr->cache.data == NULL)
+ return -ENOMSG;
+
+ if (CIPSO_V4_OPTEXIST(skb))
+ return cipso_v4_cache_add(skb, secattr);
+
+ return -ENOMSG;
+}
+
+/*
+ * Network Stack Functions
+ */
+
+/**
+ * netlbl_socket_inet_accept - Preserve the NetLabel across an accept()
+ * @sock: the old socket
+ * @newsock: the new socket
+ *
+ * Description:
+ * For kernel threads only, clear the NetLabel from @newsock and copy the
+ * NetLabel from @sock. The caller is responsibile for locking and releasing
+ * the sockets.
+ *
+ */
+void netlbl_socket_inet_accept(struct socket *sock, struct socket *newsock)
+{
+ unsigned char *opt;
+ u32 opt_len;
+
+ /* We are only interested in preserving the NetLabel on kernel
+ threads across an accept, the LSM should handle user level
+ accept calls. */
+ if (current->mm)
+ return;
+
+ if (cipso_v4_socket_getopt(sock, &opt, &opt_len) == 0 &&
+ cipso_v4_socket_setopt(newsock, opt, opt_len) != 0)
+ printk(KERN_ERR
+ "NetLabel: failed to set the CIPSOv4 option on a kernel"
+ " socket\n");
+}
+
+/*
+ * Setup Functions
+ */
+
+/**
+ * netlbl_init - Init for the NetLabel module
+ *
+ * Description:
+ * Perform the initialization of the NetLabel module before first use.
+ *
+ */
+static int __init netlbl_init(void)
+{
+ int ret_val;
+
+ printk(KERN_INFO "NetLabel: Initializing\n");
+ printk(KERN_INFO "NetLabel: domain hash size = %u\n",
+ (1 << NETLBL_DOMHSH_BITSIZE));
+ printk(KERN_INFO "NetLabel: protocols ="
+ " UNLABELED"
+ " CIPSOv4"
+ "\n");
+
+ ret_val = netlbl_domhsh_init(NETLBL_DOMHSH_BITSIZE);
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_netlink_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_unlabel_defconf();
+ if (ret_val != 0)
+ goto init_failure;
+ printk(KERN_INFO "NetLabel: unlabeled traffic allowed by default\n");
+
+ return 0;
+
+init_failure:
+ panic("NetLabel: failed to initialize properly (%d)\n", ret_val);
+}
+
+subsys_initcall(netlbl_init);
Index: linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.c
@@ -0,0 +1,661 @@
+/*
+ * NetLabel Management Support
+ *
+ * This file defines the management functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+
+#include "netlabel_domainhash.h"
+#include "netlabel_user.h"
+#include "netlabel_mgmt.h"
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_mgmt_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_MGMT_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_mgmt_add - Handle an ADD message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ADD message and add the domains from the message
+ * to the hash table. See netlabel.h for a description of the message format.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_add(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg_ptr = netlbl_netlink_payload_data(skb);
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ u32 count;
+ struct netlbl_dom_map *entry = NULL;
+ u32 iter;
+ u32 tmp_val;
+ u32 rcu_locked = 0;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto add_failure;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ count = netlbl_getinc_u32(&msg_ptr, &msg_len);
+
+ for (iter = 0; iter < count && msg_len > 0; iter++, entry = NULL) {
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (tmp_val == 0)
+ goto add_failure;
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL) {
+ ret_val = -ENOMEM;
+ goto add_failure;
+ }
+
+ if (msg_len < netlbl_align(tmp_val))
+ goto add_failure;
+ entry->domain = kmalloc(tmp_val, GFP_KERNEL);
+ if (entry->domain == NULL) {
+ ret_val = -ENOMEM;
+ goto add_failure;
+ }
+ strncpy(entry->domain, msg_ptr, tmp_val);
+ entry->domain[tmp_val - 1] = '\0';
+ msg_ptr += netlbl_align(tmp_val);
+ msg_len -= netlbl_align(tmp_val);
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ entry->type = tmp_val;
+ switch (tmp_val) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ /* XXX - we should be holding a rcu_read_lock() here
+ while we hold the result but since the entry
+ will always be deleted when the CIPSO DOI
+ is deleted we aren't going to keep the lock */
+ rcu_read_lock();
+ entry->type_def.cipsov4 = cipso_v4_doi_getdef(tmp_val);
+ if (entry->type_def.cipsov4 == NULL) {
+ rcu_read_unlock();
+ goto add_failure;
+ }
+ rcu_locked = 1;
+ break;
+ default:
+ goto add_failure;
+ }
+
+ ret_val = netlbl_domhsh_add(entry);
+ if (rcu_locked) {
+ rcu_read_unlock();
+ rcu_locked = 0;
+ }
+ if (ret_val != 0)
+ goto add_failure;
+ }
+
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+
+add_failure:
+ if (entry)
+ kfree(entry->domain);
+ kfree(entry);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_remove - Handle a REMOVE message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVE message and remove the specified domain
+ * mappings. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_remove(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg_ptr = netlbl_netlink_payload_data(skb);
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ u32 count;
+ u32 iter;
+ u32 tmp_val;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto remove_return;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto remove_return;
+ count = netlbl_getinc_u32(&msg_ptr, &msg_len);
+
+ for (iter = 0; iter < count && msg_len > 0; iter++) {
+ if (msg_len < NETLBL_LEN_U32)
+ goto remove_return;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (tmp_val == 0 ||
+ msg_len < tmp_val || msg_ptr[tmp_val - 1] != '\0')
+ goto remove_return;
+ ret_val = netlbl_domhsh_remove(msg_ptr);
+ msg_ptr += netlbl_align(tmp_val);
+ msg_len -= netlbl_align(tmp_val);
+ if (ret_val != 0)
+ goto remove_return;
+ }
+
+ ret_val = 0;
+
+remove_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and dumps the domain hash table in a
+ * form suitable for use in a kernel generated LIST message. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ struct sk_buff *ans_skb;
+
+ ans_skb = netlbl_domhsh_dump(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL)
+ goto list_failure;
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_LIST,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_adddef - Handle an ADDDEF message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ADDDEF message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_adddef(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg_ptr = netlbl_netlink_payload_data(skb);
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ struct netlbl_dom_map *entry = NULL;
+ u32 tmp_val;
+ u32 rcu_locked = 0;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto adddef_failure;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto adddef_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL) {
+ ret_val = -ENOMEM;
+ goto adddef_failure;
+ }
+
+ entry->type = tmp_val;
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (msg_len < NETLBL_LEN_U32)
+ goto adddef_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ /* XXX - we should be holding a rcu_read_lock here while we
+ hold the result but since the entry will always be
+ deleted when the CIPSO DOI is deleted we are going
+ to skip the lock */
+ rcu_read_lock();
+ entry->type_def.cipsov4 = cipso_v4_doi_getdef(tmp_val);
+ if (entry->type_def.cipsov4 == NULL) {
+ rcu_read_unlock();
+ goto adddef_failure;
+ }
+ rcu_locked = 1;
+ break;
+ default:
+ ret_val = -EINVAL;
+ goto adddef_failure;
+ }
+
+ ret_val = netlbl_domhsh_add_default(entry);
+ if (rcu_locked) {
+ rcu_read_unlock();
+ rcu_locked = 0;
+ }
+ if (ret_val != 0)
+ goto adddef_failure;
+
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+
+adddef_failure:
+ kfree(entry);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_removedef - Handle a REMOVEDEF message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVEDEF message and remove the default domain
+ * mapping. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_removedef(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto removedef_return;
+
+ ret_val = netlbl_domhsh_remove_default();
+
+removedef_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_listdef - Handle a LISTDEF message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LISTDEF message and dumps the default domain
+ * mapping in a form suitable for use in a kernel generated LISTDEF message.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_listdef(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ struct sk_buff *ans_skb;
+
+ ans_skb = netlbl_domhsh_dump_default(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL)
+ goto listdef_failure;
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_LISTDEF,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto listdef_failure;
+
+ return 0;
+
+listdef_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_modules - Handle a MODULES message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated MODULES message and respond accordingly.
+ *
+ */
+static int netlbl_mgmt_modules(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ u32 mod_count;
+ size_t data_size;
+ struct sk_buff *ans_skb = NULL;
+ unsigned char *data;
+
+ /* unlabeled + cipsov4 */
+ mod_count = 2;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32 + mod_count * NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto modules_failure;
+
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_MODULES,
+ data_size);
+ if (data == NULL)
+ goto modules_failure;
+
+ netlbl_putinc_u32(&data, mod_count, NULL);
+ netlbl_putinc_u32(&data, NETLBL_NLTYPE_UNLABELED, NULL);
+ netlbl_putinc_u32(&data, NETLBL_NLTYPE_CIPSOV4, NULL);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto modules_failure;
+
+ return 0;
+
+modules_failure:
+ kfree_skb(ans_skb);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_version - Handle a VERSION message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated VERSION message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_version(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ size_t data_size;
+ struct sk_buff *ans_skb = NULL;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto version_failure;
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_VERSION,
+ data_size);
+ if (data == NULL)
+ goto version_failure;
+
+ netlbl_put_u32(data, NETLBL_PROTO_VERSION);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto version_failure;
+
+ return 0;
+
+version_failure:
+ kfree_skb(ans_skb);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_mgmt_genl_c_add = {
+ .cmd = NLBL_MGMT_C_ADD,
+ .flags = 0,
+ .doit = netlbl_mgmt_add,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_remove = {
+ .cmd = NLBL_MGMT_C_REMOVE,
+ .flags = 0,
+ .doit = netlbl_mgmt_remove,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_list = {
+ .cmd = NLBL_MGMT_C_LIST,
+ .flags = 0,
+ .doit = netlbl_mgmt_list,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_adddef = {
+ .cmd = NLBL_MGMT_C_ADDDEF,
+ .flags = 0,
+ .doit = netlbl_mgmt_adddef,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_removedef = {
+ .cmd = NLBL_MGMT_C_REMOVEDEF,
+ .flags = 0,
+ .doit = netlbl_mgmt_removedef,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_listdef = {
+ .cmd = NLBL_MGMT_C_LISTDEF,
+ .flags = 0,
+ .doit = netlbl_mgmt_listdef,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_modules = {
+ .cmd = NLBL_MGMT_C_MODULES,
+ .flags = 0,
+ .doit = netlbl_mgmt_modules,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_version = {
+ .cmd = NLBL_MGMT_C_VERSION,
+ .flags = 0,
+ .doit = netlbl_mgmt_version,
+ .dumpit = NULL,
+};
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_mgmt_genl_init - Register the NetLabel management component
+ *
+ * Description:
+ * Register the NetLabel management component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_mgmt_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_mgmt_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_add);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_remove);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_list);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_adddef);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_removedef);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_listdef);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_modules);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_version);
+ if (ret_val != 0)
+ goto genl_init_failure;
+
+ return 0;
+
+genl_init_failure:
+ netlbl_mgmt_genl_exit();
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_genl_exit - Unregister the NetLabel management component
+ *
+ * Description:
+ * Unregister the NetLabel management component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+void netlbl_mgmt_genl_exit(void)
+{
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_version);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_modules);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_listdef);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_removedef);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_adddef);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_list);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_remove);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_add);
+
+ genl_unregister_family(&netlbl_mgmt_gnl_family);
+ netlbl_mgmt_gnl_family.id = GENL_ID_GENERATE;
+}
+
Index: linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.h
@@ -0,0 +1,250 @@
+/*
+ * NetLabel Management Support
+ *
+ * This file defines the management functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_MGMT_H
+#define _NETLABEL_MGMT_H
+
+#include <net/netlabel.h>
+
+/*
+ * The following NetLabel payloads are supported by the management interface,
+ * all of which are preceeded by the nlmsghdr struct.
+ *
+ * o ACK:
+ * Sent by the kernel in response to an applications message, applications
+ * should never send this message.
+ *
+ * +----------------------+-----------------------+
+ * | seq number (32 bits) | return code (32 bits) |
+ * +----------------------+-----------------------+
+ *
+ * seq number: the sequence number of the original message, taken from the
+ * nlmsghdr structure
+ * return code: return value, based on errno values
+ *
+ * o ADD:
+ * Sent by an application to add a domain mapping to the NetLabel system.
+ * The kernel should respond with an ACK.
+ *
+ * +-------------------+
+ * | domains (32 bits) | ...
+ * +-------------------+
+ *
+ * domains: the number of domains in the message
+ *
+ * +-------------------------+--------------------------+
+ * | domain length (32 bits) | domain string (variable) | ...
+ * +-------------------------+--------------------------+
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * domain length: the length of the "domain string" in bytes
+ * domain string: the domain string, NULL terminated
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +---------------+
+ * | doi (32 bits) |
+ * +---------------+
+ *
+ * doi: the CIPSO DOI value
+ *
+ * o REMOVE:
+ * Sent by an application to remove a domain mapping from the NetLabel
+ * system. The kernel should ACK this message.
+ *
+ * +-------------------+
+ * | domains (32 bits) | ...
+ * +-------------------+
+ *
+ * domains: the number of domains in the message
+ *
+ * +-------------------------+--------------------------+
+ * | domain length (32 bits) | domain string (variable) | ...
+ * +-------------------------+--------------------------+
+ *
+ * domain length: the length of the "domain string" in bytes
+ * domain string: the domain string, NULL terminated
+ *
+ * o LIST:
+ * This message can be sent either from an application or by the kernel in
+ * response to an application generated LIST message. When sent by an
+ * application there is no payload. The kernel should respond to a LIST
+ * message either with a LIST message on success or an ACK message on
+ * failure.
+ *
+ * +-------------------+
+ * | domains (32 bits) | ...
+ * +-------------------+
+ *
+ * domains: the number of domains in the message
+ *
+ * +-------------------------+--------------------------+
+ * | domain length (32 bits) | domain string (variable) | ...
+ * +-------------------------+--------------------------+
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * domain length: the length of the "domain string" in bytes
+ * domain string: the domain string, NULL terminated
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +----------------+---------------+
+ * | type (32 bits) | doi (32 bits) |
+ * +----------------+---------------+
+ *
+ * type: the CIPSO mapping table type (defined in the cipso_ipv4.h header
+ * as CIPSO_V4_MAP_*)
+ * doi: the CIPSO DOI value
+ *
+ * o ADDDEF:
+ * Sent by an application to set the default domain mapping for the NetLabel
+ * system. The kernel should respond with an ACK.
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +---------------+
+ * | doi (32 bits) |
+ * +---------------+
+ *
+ * doi: the CIPSO DOI value
+ *
+ * o REMOVEDEF:
+ * Sent by an application to remove the default domain mapping from the
+ * NetLabel system, there is no payload. The kernel should ACK this message.
+ *
+ * o LISTDEF:
+ * This message can be sent either from an application or by the kernel in
+ * response to an application generated LISTDEF message. When sent by an
+ * application there is no payload. The kernel should respond to a
+ * LISTDEF message either with a LISTDEF message on success or an ACK message
+ * on failure.
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +----------------+---------------+
+ * | type (32 bits) | doi (32 bits) |
+ * +----------------+---------------+
+ *
+ * type: the CIPSO mapping table type (defined in the cipso_ipv4.h header
+ * as CIPSO_V4_MAP_*)
+ * doi: the CIPSO DOI value
+ *
+ * o MODULES:
+ * Sent by an application to request a list of configured NetLabel modules
+ * in the kernel. When sent by an application there is no payload.
+ *
+ * +-------------------+
+ * | modules (32 bits) | ...
+ * +-------------------+
+ *
+ * modules: the number of modules in the message, if this is an application
+ * generated message and the value is zero then return a list of
+ * the configured modules
+ *
+ * +------------------+
+ * | module (32 bits) | ... repeated
+ * +------------------+
+ *
+ * module: the module number as defined by NETLBL_NLTYPE_*
+ *
+ * o VERSION:
+ * Sent by an application to request the NetLabel version string. When sent
+ * by an application there is no payload. This message type is also used by
+ * the kernel to respond to an VERSION request.
+ *
+ * +-------------------+
+ * | version (32 bits) |
+ * +-------------------+
+ *
+ * version: the protocol version number
+ *
+ */
+
+/* NetLabel Management commands */
+enum {
+ NLBL_MGMT_C_UNSPEC,
+ NLBL_MGMT_C_ACK,
+ NLBL_MGMT_C_ADD,
+ NLBL_MGMT_C_REMOVE,
+ NLBL_MGMT_C_LIST,
+ NLBL_MGMT_C_ADDDEF,
+ NLBL_MGMT_C_REMOVEDEF,
+ NLBL_MGMT_C_LISTDEF,
+ NLBL_MGMT_C_MODULES,
+ NLBL_MGMT_C_VERSION,
+ __NLBL_MGMT_C_MAX,
+};
+#define NLBL_MGMT_C_MAX (__NLBL_MGMT_C_MAX - 1)
+
+/* NetLabel protocol functions */
+int netlbl_mgmt_genl_init(void);
+void netlbl_mgmt_genl_exit(void);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.h
@@ -0,0 +1,99 @@
+/*
+ * NetLabel Unlabeled Support
+ *
+ * This file defines functions for dealing with unlabeled packets for the
+ * NetLabel system. The NetLabel system manages static and dynamic label
+ * mappings for network protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_UNLABELED_H
+#define _NETLABEL_UNLABELED_H
+
+#include <net/netlabel.h>
+
+/*
+ * The following NetLabel payloads are supported by the Unlabeled subsystem.
+ *
+ * o ACK:
+ * Sent by the kernel in response to an applications message, applications
+ * should never send this message.
+ *
+ * +----------------------+-----------------------+
+ * | seq number (32 bits) | return code (32 bits) |
+ * +----------------------+-----------------------+
+ *
+ * seq number: the sequence number of the original message, taken from the
+ * nlmsghdr structure
+ * return code: return value, based on errno values
+ *
+ * o ACCEPT
+ * This message is sent from an application to specify if the kernel should
+ * allow unlabled packets to pass if they do not match any of the static
+ * mappings defined in the unlabeled module.
+ *
+ * +-----------------+
+ * | allow (32 bits) |
+ * +-----------------+
+ *
+ * allow: if true (1) then allow the packets to pass, if false (0) then
+ * reject the packets
+ *
+ * o LIST
+ * This message can be sent either from an application or by the kernel in
+ * response to an application generated LIST message. When sent by an
+ * application there is no payload. The kernel should respond to a LIST
+ * message either with a LIST message on success or an ACK message on
+ * failure.
+ *
+ * +-----------------------+
+ * | accept flag (32 bits) |
+ * +-----------------------+
+ *
+ * accept flag: if true (1) then unlabeled packets are allowed to pass,
+ * if false (0) then unlabeled packets are rejected
+ *
+ */
+
+/* NetLabel Unlabeled commands */
+enum {
+ NLBL_UNLABEL_C_UNSPEC,
+ NLBL_UNLABEL_C_ACK,
+ NLBL_UNLABEL_C_ACCEPT,
+ NLBL_UNLABEL_C_LIST,
+ __NLBL_UNLABEL_C_MAX,
+};
+#define NLBL_UNLABEL_C_MAX (__NLBL_UNLABEL_C_MAX - 1)
+
+/* NetLabel protocol functions */
+int netlbl_unlabel_genl_init(void);
+void netlbl_unlabel_genl_exit(void);
+
+/* Process Unlabeled incoming network packets */
+int netlbl_unlabel_getattr(struct netlbl_lsm_secattr *secattr);
+
+/* Set the default configuration to allow Unlabeled packets */
+int netlbl_unlabel_defconf(void);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_user.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_user.c
@@ -0,0 +1,181 @@
+/*
+ * NetLabel NETLINK Interface
+ *
+ * This file defines the NETLINK interface for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/socket.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <asm/bug.h>
+
+#include "netlabel_mgmt.h"
+#include "netlabel_unlabeled.h"
+#include "netlabel_cipso_v4.h"
+#include "netlabel_user.h"
+
+/*
+ * NetLabel NETLINK Setup Functions
+ */
+
+/**
+ * netlbl_netlink_init - Initialize the NETLINK communication channel
+ *
+ * Description:
+ * Call out to the NetLabel components so they can register their families and
+ * commands with the Generic NETLINK mechanism. Returns zero on success and
+ * non-zero on failure.
+ *
+ */
+int netlbl_netlink_init(void)
+{
+ int ret_val;
+
+ ret_val = netlbl_mgmt_genl_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_cipsov4_genl_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_unlabel_genl_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ return 0;
+
+init_failure:
+ netlbl_unlabel_genl_exit();
+ return ret_val;
+}
+
+/**
+ * netlbl_exit - Shutdown the NETLINK communication channel
+ *
+ * Description:
+ * Call out to the NetLabel components so they can deregister themselves from
+ * the Generic NETLINK mechanism.
+ *
+ */
+void netlbl_netlink_exit(void)
+{
+ netlbl_unlabel_genl_exit();
+ netlbl_cipsov4_genl_exit();
+ netlbl_mgmt_genl_exit();
+}
+
+/*
+ * NetLabel Common Protocol Functions
+ */
+
+/**
+ * netlbl_netlink_send_ack - Send an ACK message
+ * @info: the generic NETLINK information
+ * @genl_family: the generic NETLINK family ID value
+ * @ack_cmd: the generic NETLINK family ACK command value
+ * @ret_code: return code to use
+ *
+ * Description:
+ * This function sends an ACK message to the sender of the NETLINK message
+ * specified by @info.
+ *
+ */
+void netlbl_netlink_send_ack(const struct genl_info *info,
+ const u32 genl_family,
+ const u8 ack_cmd,
+ const u32 ret_code)
+{
+ size_t data_size;
+ struct sk_buff *skb;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + 2 * NETLBL_LEN_U32;
+ skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (skb == NULL)
+ return;
+
+ data = netlbl_netlink_hdr_put(skb,
+ info->snd_pid,
+ 0,
+ 0,
+ genl_family,
+ ack_cmd,
+ data_size);
+ if (data == NULL)
+ goto send_ack_failure;
+
+ netlbl_putinc_u32(&data, info->snd_seq, NULL);
+ netlbl_putinc_u32(&data, ret_code, NULL);
+
+ netlbl_netlink_snd(skb, info->snd_pid);
+ return;
+
+send_ack_failure:
+ kfree_skb(skb);
+}
+
+/*
+ * NETLINK I/O Functions
+ */
+
+/**
+ * netlbl_netlink_snd - Send a NetLabel message
+ * @skb: NetLabel message
+ * @pid: destination PID
+ *
+ * Description:
+ * Sends a unicast NetLabel message over the NETLINK socket.
+ *
+ */
+int netlbl_netlink_snd(struct sk_buff *skb, const u32 pid)
+{
+ return genlmsg_unicast(skb, pid);
+}
+
+/**
+ * netlbl_netlink_snd - Send a NetLabel message
+ * @skb: NetLabel message
+ * @pid: sending PID
+ * @group: multicast group id
+ *
+ * Description:
+ * Sends a multicast NetLabel message over the NETLINK socket to all members
+ * of @group except @pid.
+ *
+ */
+int netlbl_netlink_snd_multicast(struct sk_buff *skb,
+ const u32 pid,
+ const u32 group)
+{
+ return genlmsg_multicast(skb, pid, group);
+}
Index: linux-2.6.18-rc1/net/netlabel/netlabel_user.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_user.h
@@ -0,0 +1,210 @@
+/*
+ * NetLabel NETLINK Interface
+ *
+ * This file defines the NETLINK interface for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_USER_H
+#define _NETLABEL_USER_H
+
+#include <linux/skbuff.h>
+#include <linux/capability.h>
+#include <linux/genetlink.h>
+#include <net/netlabel.h>
+#include <net/genetlink.h>
+
+/* NetLabel NETLINK helper functions */
+
+/**
+ * netlbl_netlink_cap_check - Check the NETLINK msg capabilities
+ * @skb: the NETLINK buffer
+ * @req_cap: the required capability
+ *
+ * Description:
+ * Check the NETLINK buffer's capabilities against the required capabilities.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static inline int netlbl_netlink_cap_check(const struct sk_buff *skb,
+ const kernel_cap_t req_cap)
+{
+ if (cap_raised(NETLINK_CB(skb).eff_cap, req_cap))
+ return 0;
+ return -EPERM;
+}
+
+/**
+ * netlbl_netlink_hdr_put - Write the NETLINK buffers into a buffer
+ * @skb: the buffer
+ * @pid: the PID of the receipient
+ * @seq: the sequence number
+ * @flags: NETLINK message flags
+ * @type: the generic NETLINK message family type
+ * @cmd: command
+ * @len: length
+ *
+ * Description:
+ * Write both a NETLINK nlmsghdr structure and a Generic NETLINK genlmsghdr
+ * struct to a buffer. Returns a pointer to the start of the payload buffer
+ * on success or NULL on failure.
+ *
+ */
+static inline void __netlbl_netlink_hdr(unsigned char *buffer,
+ const u32 pid,
+ const u32 seq,
+ const int flags,
+ const int type,
+ const u8 cmd,
+ const int len)
+{
+ struct nlmsghdr *nlh;
+ struct genlmsghdr *hdr;
+
+ nlh = (struct nlmsghdr *)buffer;
+ nlh->nlmsg_type = type;
+ nlh->nlmsg_len = len;
+ nlh->nlmsg_flags = flags;
+ nlh->nlmsg_pid = pid;
+ nlh->nlmsg_seq = seq;
+
+ hdr = nlmsg_data(nlh);
+ hdr->cmd = cmd;
+ hdr->version = NETLBL_PROTO_VERSION;
+ hdr->reserved = 0;
+}
+
+/**
+ * netlbl_netlink_hdr_put - Write the NETLINK buffers into a sk_buff
+ * @skb: the packet
+ * @pid: the PID of the receipient
+ * @seq: the sequence number
+ * @flags: NETLINK message flags
+ * @type: the generic NETLINK message family type
+ * @cmd: command
+ * @len: length
+ *
+ * Description:
+ * Write both a NETLINK nlmsghdr structure and a Generic NETLINK genlmsghdr
+ * struct to the packet. Returns a pointer to the start of the payload buffer
+ * on success or NULL on failure.
+ *
+ */
+static inline unsigned char *netlbl_netlink_hdr_put(struct sk_buff *skb,
+ const u32 pid,
+ const u32 seq,
+ const int flags,
+ const int type,
+ const u8 cmd,
+ const int len)
+{
+ struct nlmsghdr *nlh;
+ struct genlmsghdr *hdr;
+
+ nlh = nlmsg_put(skb, pid, seq, type, len, flags);
+ if (nlh == NULL)
+ return NULL;
+
+ hdr = nlmsg_data(nlh);
+ hdr->cmd = cmd;
+ hdr->version = NETLBL_PROTO_VERSION;
+ hdr->reserved = 0;
+
+ return (unsigned char *)hdr + GENL_HDRLEN;
+}
+
+/**
+ * netlbl_netlink_hdr_push - Write the NETLINK buffers into a sk_buff
+ * @skb: the packet
+ * @pid: the PID of the receipient
+ * @seq: the sequence number
+ * @flags: NETLINK message flags
+ * @type: the generic NETLINK message family type
+ * @cmd: command
+ * @len: length
+ *
+ * Description:
+ * Write both a NETLINK nlmsghdr structure and a Generic NETLINK genlmsghdr
+ * struct to the packet.
+ *
+ */
+static inline void netlbl_netlink_hdr_push(struct sk_buff *skb,
+ const u32 pid,
+ const u32 seq,
+ const int flags,
+ const int type,
+ const u8 cmd,
+ const int len)
+
+{
+ unsigned char *buf = skb_push(skb, NLMSG_SPACE(GENL_HDRLEN));
+ __netlbl_netlink_hdr(buf, pid, seq, flags, type, cmd, len);
+}
+
+/**
+ * netlbl_netlink_payload_len - Return the length of the payload
+ * @skb: the NETLINK buffer
+ *
+ * Description:
+ * This function returns the length of the NetLabel payload.
+ *
+ */
+static inline u32 netlbl_netlink_payload_len(const struct sk_buff *skb)
+{
+ return nlmsg_len((struct nlmsghdr *)skb->data) - GENL_HDRLEN;
+}
+
+/**
+ * netlbl_netlink_payload_data - Returns a pointer to the start of the payload
+ * @skb: the NETLINK buffer
+ *
+ * Description:
+ * This function returns a pointer to the start of the NetLabel payload.
+ *
+ */
+static inline unsigned char *netlbl_netlink_payload_data(
+ const struct sk_buff *skb)
+{
+ return nlmsg_data((struct nlmsghdr *)skb->data) + GENL_HDRLEN;
+}
+
+/* NetLabel common protocol functions */
+
+void netlbl_netlink_send_ack(const struct genl_info *info,
+ const u32 genl_family,
+ const u8 ack_cmd,
+ const u32 ret_code);
+
+/* NetLabel NETLINK I/O functions */
+
+int netlbl_netlink_init(void);
+void netlbl_netlink_exit(void);
+int netlbl_netlink_snd(struct sk_buff *skb, const u32 pid);
+int netlbl_netlink_snd_multicast(struct sk_buff *skb,
+ const u32 pid,
+ const u32 group);
+
+#endif
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 4/7] NetLabel: core NetLabel subsystem
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Add a new kernel subsystem, NetLabel, to provide explicit packet labeling
services (CIPSO, RIPSO, etc.) to LSM developers. NetLabel is designed to work
in conjunction with a LSM to intercept and decode security labels on incoming
network packets as well as ensure that outgoing network packets are labeled
according to the security mechanism employed by the LSM. The NetLabel
subsystem is configured through a Generic NETLINK interface described in the
header files included in this patch.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/Makefile | 1
net/netlabel/Kconfig | 14
net/netlabel/Makefile | 16
net/netlabel/netlabel_cipso_v4.h | 218 ++++++++++++
net/netlabel/netlabel_domainhash.c | 560 +++++++++++++++++++++++++++++++
net/netlabel/netlabel_domainhash.h | 64 +++
net/netlabel/netlabel_kapi.c | 299 ++++++++++++++++
net/netlabel/netlabel_mgmt.c | 661 +++++++++++++++++++++++++++++++++++++
net/netlabel/netlabel_mgmt.h | 250 +++++++++++++
net/netlabel/netlabel_unlabeled.h | 99 +++++
net/netlabel/netlabel_user.c | 181 ++++++++++
net/netlabel/netlabel_user.h | 210 +++++++++++
12 files changed, 2573 insertions(+)
Index: linux-2.6.18-rc1/net/Makefile
===================================================================
--- linux-2.6.18-rc1.orig/net/Makefile
+++ linux-2.6.18-rc1/net/Makefile
@@ -46,6 +46,7 @@ obj-$(CONFIG_IP_DCCP) += dccp/
obj-$(CONFIG_IP_SCTP) += sctp/
obj-$(CONFIG_IEEE80211) += ieee80211/
obj-$(CONFIG_TIPC) += tipc/
+obj-$(CONFIG_NETLABEL) += netlabel/
ifeq ($(CONFIG_NET),y)
obj-$(CONFIG_SYSCTL) += sysctl_net.o
Index: linux-2.6.18-rc1/net/netlabel/Kconfig
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/Kconfig
@@ -0,0 +1,14 @@
+#
+# NetLabel configuration
+#
+
+config NETLABEL
+ bool "NetLabel subsystem support"
+ depends on NET && SECURITY
+ default n
+ ---help---
+ NetLabel provides support for explicit network packet labeling
+ protocols such as CIPSO and RIPSO. For more information see
+ Documentation/netlabel.
+
+ If you are unsure, say N.
Index: linux-2.6.18-rc1/net/netlabel/Makefile
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/Makefile
@@ -0,0 +1,16 @@
+#
+# Makefile for the NetLabel subsystem.
+#
+# Feb 9, 2006, Paul Moore <paul.moore@hp.com>
+#
+
+# base objects
+obj-y := netlabel_user.o netlabel_kapi.o netlabel_domainhash.o
+
+# management objects
+obj-y += netlabel_mgmt.o
+
+# protocol modules
+obj-y += netlabel_unlabeled.o
+obj-y += netlabel_cipso_v4.o
+
Index: linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.h
@@ -0,0 +1,218 @@
+/*
+ * NetLabel CIPSO/IPv4 Support
+ *
+ * This file defines the CIPSO/IPv4 functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_CIPSO_V4
+#define _NETLABEL_CIPSO_V4
+
+#include <net/netlabel.h>
+
+/*
+ * The following NetLabel payloads are supported by the CIPSO subsystem, all
+ * of which are preceeded by the nlmsghdr struct.
+ *
+ * o ACK:
+ * Sent by the kernel in response to an applications message, applications
+ * should never send this message.
+ *
+ * +----------------------+-----------------------+
+ * | seq number (32 bits) | return code (32 bits) |
+ * +----------------------+-----------------------+
+ *
+ * seq number: the sequence number of the original message, taken from the
+ * nlmsghdr structure
+ * return code: return value, based on errno values
+ *
+ * o ADD:
+ * Sent by an application to add a new DOI mapping table, after completion
+ * of the task the kernel should ACK this message.
+ *
+ * +---------------+--------------------+---------------------+
+ * | DOI (32 bits) | map type (32 bits) | tag count (32 bits) | ...
+ * +---------------+--------------------+---------------------+
+ *
+ * +-----------------+
+ * | tag #X (8 bits) | ... repeated
+ * +-----------------+
+ *
+ * +-------------- ---- --- -- -
+ * | mapping data
+ * +-------------- ---- --- -- -
+ *
+ * DOI: the DOI value
+ * map type: the mapping table type (defined in the cipso_ipv4.h header
+ * as CIPSO_V4_MAP_*)
+ * tag count: the number of tags, must be greater than zero
+ * tag: the CIPSO tag for the DOI, tags listed first are given
+ * higher priorirty when sending packets
+ * mapping data: specific to the map type (see below)
+ *
+ * CIPSO_V4_MAP_STD
+ *
+ * +------------------+-----------------------+----------------------+
+ * | levels (32 bits) | max l level (32 bits) | max r level (8 bits) | ...
+ * +------------------+-----------------------+----------------------+
+ *
+ * +----------------------+---------------------+---------------------+
+ * | categories (32 bits) | max l cat (32 bits) | max r cat (16 bits) | ...
+ * +----------------------+---------------------+---------------------+
+ *
+ * +--------------------------+-------------------------+
+ * | local level #X (32 bits) | CIPSO level #X (8 bits) | ... repeated
+ * +--------------------------+-------------------------+
+ *
+ * +-----------------------------+-----------------------------+
+ * | local category #X (32 bits) | CIPSO category #X (16 bits) | ... repeated
+ * +-----------------------------+-----------------------------+
+ *
+ * levels: the number of level mappings
+ * max l level: the highest local level
+ * max r level: the highest remote/CIPSO level
+ * categories: the number of category mappings
+ * max l cat: the highest local category
+ * max r cat: the highest remote/CIPSO category
+ * local level: the local part of a level mapping
+ * CIPSO level: the remote/CIPSO part of a level mapping
+ * local category: the local part of a category mapping
+ * CIPSO category: the remote/CIPSO part of a category mapping
+ *
+ * CIPSO_V4_MAP_PASS
+ *
+ * No mapping data is needed for this map type.
+ *
+ * o REMOVE:
+ * Sent by an application to remove a specific DOI mapping table from the
+ * CIPSO V4 system. The kernel should ACK this message.
+ *
+ * +---------------+
+ * | DOI (32 bits) |
+ * +---------------+
+ *
+ * DOI: the DOI value
+ *
+ * o LIST:
+ * Sent by an application to list the details of a DOI definition. The
+ * kernel should send an ACK on error or a response as indicated below. The
+ * application generated message format is shown below.
+ *
+ * +---------------+
+ * | DOI (32 bits) |
+ * +---------------+
+ *
+ * DOI: the DOI value
+ *
+ * The valid response message format depends on the type of the DOI mapping,
+ * the known formats are shown below.
+ *
+ * +--------------------+
+ * | map type (32 bits) | ...
+ * +--------------------+
+ *
+ * map type: the DOI mapping table type (defined in the cipso_ipv4.h
+ * header as CIPSO_V4_MAP_*)
+ *
+ * (map type == CIPSO_V4_MAP_STD)
+ *
+ * +----------------+------------------+----------------------+
+ * | tags (32 bits) | levels (32 bits) | categories (32 bits) | ...
+ * +----------------+------------------+----------------------+
+ *
+ * +-----------------+
+ * | tag #X (8 bits) | ... repeated
+ * +-----------------+
+ *
+ * +--------------------------+-------------------------+
+ * | local level #X (32 bits) | CIPSO level #X (8 bits) | ... repeated
+ * +--------------------------+-------------------------+
+ *
+ * +-----------------------------+-----------------------------+
+ * | local category #X (32 bits) | CIPSO category #X (16 bits) | ... repeated
+ * +-----------------------------+-----------------------------+
+ *
+ * tags: the number of CIPSO tag types
+ * levels: the number of level mappings
+ * categories: the number of category mappings
+ * tag: the tag number, tags listed first are given higher
+ * priority when sending packets
+ * local level: the local part of a level mapping
+ * CIPSO level: the remote/CIPSO part of a level mapping
+ * local category: the local part of a category mapping
+ * CIPSO category: the remote/CIPSO part of a category mapping
+ *
+ * (map type == CIPSO_V4_MAP_PASS)
+ *
+ * +----------------+
+ * | tags (32 bits) | ...
+ * +----------------+
+ *
+ * +-----------------+
+ * | tag #X (8 bits) | ... repeated
+ * +-----------------+
+ *
+ * tags: the number of CIPSO tag types
+ * tag: the tag number, tags listed first are given higher
+ * priority when sending packets
+ *
+ * o LISTALL:
+ * This message is sent by an application to list the valid DOIs on the
+ * system. There is no payload and the kernel should respond with an ACK
+ * or the following message.
+ *
+ * +---------------------+------------------+-----------------------+
+ * | DOI count (32 bits) | DOI #X (32 bits) | map type #X (32 bits) |
+ * +---------------------+------------------+-----------------------+
+ *
+ * +-----------------------+
+ * | map type #X (32 bits) | ...
+ * +-----------------------+
+ *
+ * DOI count: the number of DOIs
+ * DOI: the DOI value
+ * map type: the DOI mapping table type (defined in the cipso_ipv4.h
+ * header as CIPSO_V4_MAP_*)
+ *
+ */
+
+/* NetLabel CIPSOv4 commands */
+enum {
+ NLBL_CIPSOV4_C_UNSPEC,
+ NLBL_CIPSOV4_C_ACK,
+ NLBL_CIPSOV4_C_ADD,
+ NLBL_CIPSOV4_C_REMOVE,
+ NLBL_CIPSOV4_C_LIST,
+ NLBL_CIPSOV4_C_LISTALL,
+ __NLBL_CIPSOV4_C_MAX,
+};
+#define NLBL_CIPSOV4_C_MAX (__NLBL_CIPSOV4_C_MAX - 1)
+
+/* NetLabel protocol functions */
+int netlbl_cipsov4_genl_init(void);
+void netlbl_cipsov4_genl_exit(void);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.c
@@ -0,0 +1,560 @@
+/*
+ * NetLabel Domain Hash Table
+ *
+ * This file manages the domain hash table that NetLabel uses to determine
+ * which network labeling protocol to use for a given domain. The NetLabel
+ * system manages static and dynamic label mappings for network protocols such
+ * as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/skbuff.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+#include "netlabel_mgmt.h"
+#include "netlabel_domainhash.h"
+
+struct netlbl_domhsh_tbl {
+ struct list_head *tbl;
+ u32 size;
+};
+
+/* Domain hash table */
+/* XXX - updates should be so rare that having one spinlock for the entire
+ hash table should be okay */
+DEFINE_SPINLOCK(netlbl_domhsh_lock);
+static struct netlbl_domhsh_tbl *netlbl_domhsh = NULL;
+
+/* Default domain mapping */
+DEFINE_SPINLOCK(netlbl_domhsh_def_lock);
+static struct netlbl_dom_map *netlbl_domhsh_def = NULL;
+
+/*
+ * Domain Hash Table Helper Functions
+ */
+
+/**
+ * netlbl_domhsh_free_entry - Frees a domain hash table entry
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to a hash table entry can be released
+ * safely.
+ *
+ */
+static void netlbl_domhsh_free_entry(struct rcu_head *entry)
+{
+ struct netlbl_dom_map *ptr;
+
+ ptr = container_of(entry, struct netlbl_dom_map, rcu);
+ kfree(ptr->domain);
+ kfree(ptr);
+}
+
+/**
+ * netlbl_domhsh_hash - Hashing function for the domain hash table
+ * @domain: the domain name to hash
+ *
+ * Description:
+ * This is the hashing function for the domain hash table, it returns the
+ * correct bucket number for the domain. The caller is responsibile for
+ * calling the rcu_read_[un]lock() functions.
+ *
+ */
+static u32 netlbl_domhsh_hash(const char *key)
+{
+ u32 iter;
+ u32 val;
+ u32 len;
+
+ /* This is taken (with slight modification) from
+ security/selinux/ss/symtab.c:symhash() */
+
+ for (iter = 0, val = 0, len = strlen(key); iter < len; iter++)
+ val = (val << 4 | (val >> (8 * sizeof(u32) - 4))) ^ key[iter];
+ return val & (rcu_dereference(netlbl_domhsh)->size - 1);
+}
+
+/**
+ * netlbl_domhsh_search - Search for a domain entry
+ * @domain: the domain
+ * @def: return default if no match is found
+ *
+ * Description:
+ * Searches the domain hash table and returns a pointer to the hash table
+ * entry if found, otherwise NULL is returned. If @def is non-zero and a
+ * match is not found in the domain hash table the default mapping is returned
+ * if it exists. The caller is responsibile for the rcu hash table locks
+ * (i.e. the caller much call rcu_read_[un]lock()).
+ *
+ */
+static struct netlbl_dom_map *netlbl_domhsh_search(const char *domain,
+ const u32 def)
+{
+ u32 bkt;
+ struct netlbl_dom_map *iter;
+
+ if (domain != NULL) {
+ bkt = netlbl_domhsh_hash(domain);
+ list_for_each_entry_rcu(iter, &netlbl_domhsh->tbl[bkt], list)
+ if (iter->valid && strcmp(iter->domain, domain) == 0)
+ return iter;
+ }
+
+ if (def != 0) {
+ iter = rcu_dereference(netlbl_domhsh_def);
+ if (iter != NULL && iter->valid)
+ return iter;
+ }
+
+ return NULL;
+}
+
+/*
+ * Domain Hash Table Functions
+ */
+
+/**
+ * netlbl_domhsh_init - Init for the domain hash
+ * @size: the number of bits to use for the hash buckets
+ *
+ * Description:
+ * Initializes the domain hash table, should be called only by
+ * netlbl_user_init() during module load. Returns zero on success, non-zero
+ * values on error.
+ *
+ */
+int netlbl_domhsh_init(const u32 size)
+{
+ u32 iter;
+ struct netlbl_domhsh_tbl *hsh_tbl;
+
+ if (size == 0)
+ return -EINVAL;
+
+ hsh_tbl = kmalloc(sizeof(*hsh_tbl), GFP_KERNEL);
+ if (hsh_tbl == NULL)
+ return -ENOMEM;
+ hsh_tbl->size = 1 << size;
+ hsh_tbl->tbl = kcalloc(hsh_tbl->size,
+ sizeof(struct list_head),
+ GFP_KERNEL);
+ if (hsh_tbl->tbl == NULL) {
+ kfree(hsh_tbl);
+ return -ENOMEM;
+ }
+ for (iter = 0; iter < hsh_tbl->size; iter++)
+ INIT_LIST_HEAD(&hsh_tbl->tbl[iter]);
+
+ rcu_read_lock();
+ spin_lock(&netlbl_domhsh_lock);
+ rcu_assign_pointer(netlbl_domhsh, hsh_tbl);
+ spin_unlock(&netlbl_domhsh_lock);
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * netlbl_domhsh_exit - Cleanup for the domain hash
+ *
+ * Description:
+ * Clears and frees the domain hash table. No real serious effort is made
+ * towards locking but that shouldn't be an issue since we should only be
+ * called when the module is being unloaded. Returns zero on success, non-zero
+ * values on error.
+ *
+ */
+void netlbl_domhsh_exit(void)
+{
+ u32 iter_bkt;
+ struct netlbl_domhsh_tbl *hsh_tbl;
+ struct netlbl_dom_map *iter_list;
+
+ rcu_read_lock();
+ if (rcu_dereference(netlbl_domhsh) == NULL) {
+ rcu_read_unlock();
+ return;
+ }
+ iter_list = rcu_dereference(netlbl_domhsh_def);
+ spin_lock(&netlbl_domhsh_def_lock);
+ rcu_assign_pointer(netlbl_domhsh_def, NULL);
+ spin_unlock(&netlbl_domhsh_def_lock);
+ rcu_read_unlock();
+
+ if (iter_list) {
+ synchronize_rcu();
+ netlbl_domhsh_free_entry(&iter_list->rcu);
+ }
+
+ rcu_read_lock();
+ hsh_tbl = rcu_dereference(netlbl_domhsh);
+ spin_lock(&netlbl_domhsh_lock);
+ rcu_assign_pointer(netlbl_domhsh, NULL);
+ spin_unlock(&netlbl_domhsh_lock);
+ rcu_read_unlock();
+
+ synchronize_rcu();
+
+ /* don't use the rcu functions here since the table is no longer in use
+ outside this function's scope */
+ for (iter_bkt = 0; iter_bkt < hsh_tbl->size; iter_bkt++)
+ list_for_each_entry(iter_list, &hsh_tbl->tbl[iter_bkt], list) {
+ iter_list->valid = 0;
+ list_del(&iter_list->list);
+ netlbl_domhsh_free_entry(&iter_list->rcu);
+ }
+ kfree(hsh_tbl->tbl);
+ kfree(hsh_tbl);
+}
+
+/**
+ * netlbl_domhsh_add - Adds a entry to the domain hash table
+ * @entry: the entry to add
+ *
+ * Description:
+ * Adds a new entry to the domain hash table and handles any updates to the
+ * lower level protocol handler (i.e. CIPSO). Returns zero on success,
+ * negative on failure.
+ *
+ */
+int netlbl_domhsh_add(struct netlbl_dom_map *entry)
+{
+ int ret_val = -EPERM;
+ u32 bkt;
+
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ ret_val = 0;
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ ret_val = cipso_v4_doi_domhsh_add(entry->type_def.cipsov4,
+ entry->domain);
+ break;
+ }
+ if (ret_val != 0)
+ return ret_val;
+
+ entry->valid = 1;
+ INIT_RCU_HEAD(&entry->rcu);
+
+ rcu_read_lock();
+ if (entry->domain != NULL &&
+ netlbl_domhsh_search(entry->domain, 0) == NULL) {
+ bkt = netlbl_domhsh_hash(entry->domain);
+ spin_lock(&netlbl_domhsh_lock);
+ list_add_tail_rcu(&entry->list, &netlbl_domhsh->tbl[bkt]);
+ spin_unlock(&netlbl_domhsh_lock);
+ } else if (entry->domain == NULL &&
+ rcu_dereference(netlbl_domhsh_def) == NULL) {
+ INIT_LIST_HEAD(&entry->list);
+ spin_lock(&netlbl_domhsh_def_lock);
+ rcu_assign_pointer(netlbl_domhsh_def, entry);
+ spin_unlock(&netlbl_domhsh_def_lock);
+ } else {
+ rcu_read_unlock();
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ BUG_ON(cipso_v4_doi_domhsh_remove(
+ entry->type_def.cipsov4,
+ entry->domain));
+ break;
+ }
+ return -EPERM;
+ }
+ rcu_read_unlock();
+
+ return 0;
+}
+
+/**
+ * netlbl_domhsh_add_default - Adds the default entry to the domain hash table
+ * @entry: the entry to add
+ *
+ * Description:
+ * Adds a new default entry to the domain hash table and handles any updates
+ * to the lower level protocol handler (i.e. CIPSO). Returns zero on success,
+ * negative on failure.
+ *
+ */
+int netlbl_domhsh_add_default(struct netlbl_dom_map *entry)
+{
+ return netlbl_domhsh_add(entry);
+}
+
+/**
+ * netlbl_domhsh_remove - Removes an entry from the domain hash table
+ * @domain: the domain to remove
+ *
+ * Description:
+ * Removes an entry from the domain hash table and handles any updates to the
+ * lower level protocol handler (i.e. CIPSO). Returns zero on success,
+ * negative on failure.
+ *
+ */
+int netlbl_domhsh_remove(const char *domain)
+{
+ int ret_val = -EPERM;
+ struct netlbl_dom_map *entry;
+
+ rcu_read_lock();
+ if (domain != NULL)
+ entry = netlbl_domhsh_search(domain, 0);
+ else
+ entry = netlbl_domhsh_search(domain, 1);
+ if (entry == NULL)
+ goto remove_return;
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (cipso_v4_doi_domhsh_remove(entry->type_def.cipsov4,
+ entry->domain) != 0)
+ goto remove_return;
+ break;
+ }
+ if (entry != rcu_dereference(netlbl_domhsh_def)) {
+ spin_lock(&netlbl_domhsh_lock);
+ entry->valid = 0;
+ list_del_rcu(&entry->list);
+ spin_unlock(&netlbl_domhsh_lock);
+ } else {
+ spin_lock(&netlbl_domhsh_def_lock);
+ entry->valid = 0;
+ rcu_assign_pointer(netlbl_domhsh_def, NULL);
+ spin_unlock(&netlbl_domhsh_def_lock);
+ }
+ call_rcu(&entry->rcu, netlbl_domhsh_free_entry);
+
+ ret_val = 0;
+
+remove_return:
+ rcu_read_unlock();
+ return ret_val;
+}
+
+/**
+ * netlbl_domhsh_remove_default - Removes the default entry from the table
+ *
+ * Description:
+ * Removes/resets the default entry for the domain hash table and handles any
+ * updates to the lower level protocol handler (i.e. CIPSO). Returns zero on
+ * success, non-zero on failure.
+ *
+ */
+int netlbl_domhsh_remove_default(void)
+{
+ return netlbl_domhsh_remove(NULL);
+}
+
+/**
+ * netlbl_domhsh_getentry - Get an entry from the domain hash table
+ * @domain: the domain name to search for
+ *
+ * Description:
+ * Look through the domain hash table searching for an entry to match @domain,
+ * return a pointer to a copy of the entry or NULL. The caller is responsibile
+ * for ensuring that rcu_read_[un]lock() is called.
+ *
+ */
+struct netlbl_dom_map *netlbl_domhsh_getentry(const char *domain)
+{
+ return netlbl_domhsh_search(domain, 1);
+}
+
+/**
+ * netlbl_domhsh_dump - Dump the domain hash table into a sk_buff
+ *
+ * Description:
+ * Dump the domain hash table into a buffer suitable for returning to an
+ * application in response to a NetLabel management DOMAIN message. This
+ * function may fail if another process is growing the hash table at the same
+ * time. The returned sk_buff has room at the front of the sk_buff for
+ * @headroom bytes. See netlabel.h for the DOMAIN message format. Returns a
+ * pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *netlbl_domhsh_dump(const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ ssize_t buf_len;
+ u32 bkt_iter;
+ u32 dom_cnt = 0;
+ struct netlbl_domhsh_tbl *hsh_tbl;
+ struct netlbl_dom_map *list_iter;
+ ssize_t tmp_len;
+
+ /* XXX - This is kinda ugly as we have to go through the table once to
+ determine how large of a buffer we need, drop the locks, allocate
+ the buffer, grab the locks, and finally fill the buffer. The
+ problem is that there is that open window where the table could
+ grow and we will end up short on space. */
+
+ buf_len = NETLBL_LEN_U32;
+ rcu_read_lock();
+ hsh_tbl = rcu_dereference(netlbl_domhsh);
+ for (bkt_iter = 0; bkt_iter < hsh_tbl->size; bkt_iter++)
+ list_for_each_entry_rcu(list_iter,
+ &hsh_tbl->tbl[bkt_iter], list) {
+ buf_len += 2 * NETLBL_LEN_U32 +
+ netlbl_align(strlen(list_iter->domain) + 1);
+ switch (list_iter->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ buf_len += 2 * NETLBL_LEN_U32;
+ break;
+ }
+ dom_cnt++;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+ netlbl_putinc_u32(&buf, dom_cnt, &buf_len);
+
+ rcu_read_lock();
+ hsh_tbl = rcu_dereference(netlbl_domhsh);
+ for (bkt_iter = 0; bkt_iter < hsh_tbl->size; bkt_iter++)
+ list_for_each_entry_rcu(list_iter,
+ &hsh_tbl->tbl[bkt_iter], list) {
+ tmp_len = netlbl_align(strlen(list_iter->domain) + 1);
+ if (buf_len < 2 * NETLBL_LEN_U32 + tmp_len)
+ goto dump_failure;
+ netlbl_putinc_u32(&buf, tmp_len, &buf_len);
+ netlbl_putinc_str(&buf, list_iter->domain, &buf_len);
+ netlbl_putinc_u32(&buf, list_iter->type, &buf_len);
+ switch (list_iter->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto dump_failure;
+ netlbl_putinc_u32(
+ &buf,
+ list_iter->type_def.cipsov4->type,
+ &buf_len);
+ netlbl_putinc_u32(
+ &buf,
+ list_iter->type_def.cipsov4->doi,
+ &buf_len);
+ break;
+ }
+ }
+ rcu_read_unlock();
+
+ return skb;
+
+dump_failure:
+ rcu_read_unlock();
+ kfree_skb(skb);
+ return NULL;
+}
+
+/**
+ * netlbl_domhsh_dump_default - Dump the default domain mapping into a sk_buff
+ *
+ * Description:
+ * Dump the default domain mapping into a buffer suitable for returning to an
+ * application in response to a NetLabel management DEFDOMAIN message. This
+ * function may fail if another process is changing the default domain mapping
+ * at the same time. The returned sk_buff has room at the front of the
+ * skb_buff for @headroom bytes. See netlabel.h for the DEFDOMAIN message
+ * format. Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *netlbl_domhsh_dump_default(const size_t headroom)
+{
+ struct sk_buff *skb;
+ unsigned char *buf;
+ ssize_t buf_len;
+ struct netlbl_dom_map *entry;
+
+ /* XXX - This is kinda ugly as we have to go look at the default
+ mapping once to determine how large of a buffer we need, drop the
+ locks, allocate the buffer, grab the locks, and finally fill the
+ buffer. The problem is that there is that open window where the
+ default mapping could change on us, if that is the case we fail. */
+
+ buf_len = NETLBL_LEN_U32;
+ rcu_read_lock();
+ entry = rcu_dereference(netlbl_domhsh_def);
+ if (entry != NULL)
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ buf_len += 2 * NETLBL_LEN_U32;
+ break;
+ }
+ rcu_read_unlock();
+
+ skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
+ if (skb == NULL)
+ return NULL;
+ buf = skb_put(skb, buf_len);
+
+ rcu_read_lock();
+ if (entry != rcu_dereference(netlbl_domhsh_def))
+ goto dump_default_failure;
+ if (entry != NULL) {
+ netlbl_putinc_u32(&buf, entry->type, &buf_len);
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (buf_len < 2 * NETLBL_LEN_U32)
+ goto dump_default_failure;
+ netlbl_putinc_u32(&buf,
+ entry->type_def.cipsov4->type,
+ &buf_len);
+ netlbl_putinc_u32(&buf,
+ entry->type_def.cipsov4->doi,
+ &buf_len);
+ break;
+ }
+ } else
+ netlbl_putinc_u32(&buf, NETLBL_NLTYPE_NONE, &buf_len);
+ rcu_read_unlock();
+
+ return skb;
+
+dump_default_failure:
+ rcu_read_unlock();
+ kfree_skb(skb);
+ return NULL;
+}
Index: linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_domainhash.h
@@ -0,0 +1,64 @@
+/*
+ * NetLabel Domain Hash Table
+ *
+ * This file manages the domain hash table that NetLabel uses to determine
+ * which network labeling protocol to use for a given domain. The NetLabel
+ * system manages static and dynamic label mappings for network protocols such
+ * as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_DOMAINHASH_H
+#define _NETLABEL_DOMAINHASH_H
+
+/* Domain hash table size */
+/* XXX - currently this number is an uneducated guess */
+#define NETLBL_DOMHSH_BITSIZE 7
+
+/* Domain mapping definition struct */
+struct netlbl_dom_map {
+ char *domain;
+ u32 type;
+ union {
+ struct cipso_v4_doi *cipsov4;
+ } type_def;
+
+ u32 valid;
+ struct list_head list;
+ struct rcu_head rcu;
+};
+
+/* init/exit functions */
+int netlbl_domhsh_init(const u32 size);
+void netlbl_domhsh_exit(void);
+
+/* Manipulate the domain hash table */
+int netlbl_domhsh_add(struct netlbl_dom_map *entry);
+int netlbl_domhsh_add_default(struct netlbl_dom_map *entry);
+int netlbl_domhsh_remove_default(void);
+struct netlbl_dom_map *netlbl_domhsh_getentry(const char *domain);
+struct sk_buff *netlbl_domhsh_dump(const size_t headroom);
+struct sk_buff *netlbl_domhsh_dump_default(const size_t headroom);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_kapi.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_kapi.c
@@ -0,0 +1,299 @@
+/*
+ * NetLabel Kernel API
+ *
+ * This file defines the kernel API for the NetLabel system. The NetLabel
+ * system manages static and dynamic label mappings for network protocols such
+ * as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <net/ip.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+#include "netlabel_domainhash.h"
+#include "netlabel_unlabeled.h"
+#include "netlabel_user.h"
+
+/*
+ * LSM Functions
+ */
+
+/**
+ * netlbl_socket_setattr - Label a socket using the correct protocol
+ * @sock: the socket to label
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Attach the correct label to the given socket using the security attributes
+ * specified in @secattr. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_socket_setattr(const struct socket *sock,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val = -ENOENT;
+ struct netlbl_dom_map *dom_entry;
+
+ rcu_read_lock();
+ dom_entry = netlbl_domhsh_getentry(secattr->domain);
+ if (dom_entry == NULL)
+ goto socket_setattr_return;
+ switch (dom_entry->type) {
+ case NETLBL_NLTYPE_CIPSOV4:
+ ret_val = cipso_v4_socket_setattr(sock,
+ dom_entry->type_def.cipsov4,
+ secattr);
+ break;
+ case NETLBL_NLTYPE_UNLABELED:
+ ret_val = 0;
+ break;
+ default:
+ ret_val = -ENOENT;
+ }
+
+socket_setattr_return:
+ rcu_read_unlock();
+ return ret_val;
+}
+
+/**
+ * netlbl_socket_peekattr - Get the security attributes on a queued packet
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Peek at the incoming packet queue of @sock and return the security
+ * attributes of the first packet in the queue. If there are no packets in
+ * the queue or no packets with security attributes return -ENOMSG. Otherwise
+ * return zero on success, negative values on failure.
+ *
+ */
+int netlbl_socket_peekattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+ struct sk_buff_head *sock_queue;
+ struct sk_buff *skb;
+ struct sk_buff *skb_tmp = NULL;
+
+ sock_queue = &sock->sk->sk_receive_queue;
+ /* XXX - We could just take the lock here and do a skb_peek() but we
+ might end up holding the lock for a long time so grab the
+ lock, do a peek, and clone the result */
+ spin_lock(&sock_queue->lock);
+ skb = skb_peek(sock_queue);
+ if (skb != NULL)
+ skb_tmp = skb_clone(skb, GFP_ATOMIC);
+ spin_unlock(&sock_queue->lock);
+ if (skb_tmp != NULL) {
+ ret_val = netlbl_skbuff_getattr(skb_tmp, secattr);
+ kfree_skb(skb_tmp);
+ } else
+ ret_val = -ENOMEM;
+
+ return ret_val;
+}
+
+/**
+ * netlbl_socket_getattr - Determine the security attributes of a socket
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Examines the given socket to see any NetLabel style labeling has been
+ * applied to the socket, if so it parses the socket label and returns the
+ * security attributes in @secattr. Returns zero on success, negative values
+ * on failure.
+ *
+ */
+int netlbl_socket_getattr(const struct socket *sock,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+
+ ret_val = cipso_v4_socket_getattr(sock, secattr);
+ if (ret_val == 0)
+ return 0;
+
+ return netlbl_unlabel_getattr(secattr);
+}
+
+/**
+ * netlbl_skbuff_getattr - Determine the security attributes of a packet
+ * @skb: the packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Examines the given packet to see if a recognized form of packet labeling
+ * is present, if so it parses the packet label and returns the security
+ * attributes in @secattr. Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int netlbl_skbuff_getattr(const struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr)
+{
+ int ret_val;
+
+ ret_val = cipso_v4_skbuff_getattr(skb, secattr);
+ if (ret_val == 0)
+ return 0;
+
+ return netlbl_unlabel_getattr(secattr);
+}
+
+/**
+ * netlbl_skbuff_err - Handle a LSM error on a sk_buff
+ * @skb: the packet
+ * @error: the error code
+ *
+ * Description:
+ * Deal with a LSM problem when handling the packet in @skb, typically this is
+ * a permission denied problem (-EACCES). The correct action is determined
+ * according to the packet's labeling protocol.
+ *
+ */
+void netlbl_skbuff_err(struct sk_buff *skb, int error)
+{
+ if (CIPSO_V4_OPTEXIST(skb))
+ cipso_v4_error(skb, error, 0);
+}
+
+/**
+ * netlbl_cache_invalidate - Invalidate all of the NetLabel protocol caches
+ *
+ * Description:
+ * For all of the NetLabel protocols that support some form of label mapping
+ * cache, invalidate the cache. Returns zero on success, negative values on
+ * error.
+ *
+ */
+void netlbl_cache_invalidate(void)
+{
+ cipso_v4_cache_invalidate();
+}
+
+/**
+ * netlbl_cache_add - Add an entry to a NetLabel protocol cache
+ * @skb: the packet
+ * @secattr: the packet's security attributes
+ *
+ * Description:
+ * Add the LSM security attributes for the given packet to the underlying
+ * NetLabel protocol's label mapping cache. Returns zero on success, negative
+ * values on error.
+ *
+ */
+int netlbl_cache_add(const struct sk_buff *skb,
+ const struct netlbl_lsm_secattr *secattr)
+{
+ if (secattr->cache.data == NULL)
+ return -ENOMSG;
+
+ if (CIPSO_V4_OPTEXIST(skb))
+ return cipso_v4_cache_add(skb, secattr);
+
+ return -ENOMSG;
+}
+
+/*
+ * Network Stack Functions
+ */
+
+/**
+ * netlbl_socket_inet_accept - Preserve the NetLabel across an accept()
+ * @sock: the old socket
+ * @newsock: the new socket
+ *
+ * Description:
+ * For kernel threads only, clear the NetLabel from @newsock and copy the
+ * NetLabel from @sock. The caller is responsibile for locking and releasing
+ * the sockets.
+ *
+ */
+void netlbl_socket_inet_accept(struct socket *sock, struct socket *newsock)
+{
+ unsigned char *opt;
+ u32 opt_len;
+
+ /* We are only interested in preserving the NetLabel on kernel
+ threads across an accept, the LSM should handle user level
+ accept calls. */
+ if (current->mm)
+ return;
+
+ if (cipso_v4_socket_getopt(sock, &opt, &opt_len) == 0 &&
+ cipso_v4_socket_setopt(newsock, opt, opt_len) != 0)
+ printk(KERN_ERR
+ "NetLabel: failed to set the CIPSOv4 option on a kernel"
+ " socket\n");
+}
+
+/*
+ * Setup Functions
+ */
+
+/**
+ * netlbl_init - Init for the NetLabel module
+ *
+ * Description:
+ * Perform the initialization of the NetLabel module before first use.
+ *
+ */
+static int __init netlbl_init(void)
+{
+ int ret_val;
+
+ printk(KERN_INFO "NetLabel: Initializing\n");
+ printk(KERN_INFO "NetLabel: domain hash size = %u\n",
+ (1 << NETLBL_DOMHSH_BITSIZE));
+ printk(KERN_INFO "NetLabel: protocols ="
+ " UNLABELED"
+ " CIPSOv4"
+ "\n");
+
+ ret_val = netlbl_domhsh_init(NETLBL_DOMHSH_BITSIZE);
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_netlink_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_unlabel_defconf();
+ if (ret_val != 0)
+ goto init_failure;
+ printk(KERN_INFO "NetLabel: unlabeled traffic allowed by default\n");
+
+ return 0;
+
+init_failure:
+ panic("NetLabel: failed to initialize properly (%d)\n", ret_val);
+}
+
+subsys_initcall(netlbl_init);
Index: linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.c
@@ -0,0 +1,661 @@
+/*
+ * NetLabel Management Support
+ *
+ * This file defines the management functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+
+#include "netlabel_domainhash.h"
+#include "netlabel_user.h"
+#include "netlabel_mgmt.h"
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_mgmt_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_MGMT_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_mgmt_add - Handle an ADD message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ADD message and add the domains from the message
+ * to the hash table. See netlabel.h for a description of the message format.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_add(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg_ptr = netlbl_netlink_payload_data(skb);
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ u32 count;
+ struct netlbl_dom_map *entry = NULL;
+ u32 iter;
+ u32 tmp_val;
+ u32 rcu_locked = 0;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto add_failure;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ count = netlbl_getinc_u32(&msg_ptr, &msg_len);
+
+ for (iter = 0; iter < count && msg_len > 0; iter++, entry = NULL) {
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (tmp_val == 0)
+ goto add_failure;
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL) {
+ ret_val = -ENOMEM;
+ goto add_failure;
+ }
+
+ if (msg_len < netlbl_align(tmp_val))
+ goto add_failure;
+ entry->domain = kmalloc(tmp_val, GFP_KERNEL);
+ if (entry->domain == NULL) {
+ ret_val = -ENOMEM;
+ goto add_failure;
+ }
+ strncpy(entry->domain, msg_ptr, tmp_val);
+ entry->domain[tmp_val - 1] = '\0';
+ msg_ptr += netlbl_align(tmp_val);
+ msg_len -= netlbl_align(tmp_val);
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ entry->type = tmp_val;
+ switch (tmp_val) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ /* XXX - we should be holding a rcu_read_lock() here
+ while we hold the result but since the entry
+ will always be deleted when the CIPSO DOI
+ is deleted we aren't going to keep the lock */
+ rcu_read_lock();
+ entry->type_def.cipsov4 = cipso_v4_doi_getdef(tmp_val);
+ if (entry->type_def.cipsov4 == NULL) {
+ rcu_read_unlock();
+ goto add_failure;
+ }
+ rcu_locked = 1;
+ break;
+ default:
+ goto add_failure;
+ }
+
+ ret_val = netlbl_domhsh_add(entry);
+ if (rcu_locked) {
+ rcu_read_unlock();
+ rcu_locked = 0;
+ }
+ if (ret_val != 0)
+ goto add_failure;
+ }
+
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+
+add_failure:
+ if (entry)
+ kfree(entry->domain);
+ kfree(entry);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_remove - Handle a REMOVE message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVE message and remove the specified domain
+ * mappings. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_remove(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg_ptr = netlbl_netlink_payload_data(skb);
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ u32 count;
+ u32 iter;
+ u32 tmp_val;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto remove_return;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto remove_return;
+ count = netlbl_getinc_u32(&msg_ptr, &msg_len);
+
+ for (iter = 0; iter < count && msg_len > 0; iter++) {
+ if (msg_len < NETLBL_LEN_U32)
+ goto remove_return;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (tmp_val == 0 ||
+ msg_len < tmp_val || msg_ptr[tmp_val - 1] != '\0')
+ goto remove_return;
+ ret_val = netlbl_domhsh_remove(msg_ptr);
+ msg_ptr += netlbl_align(tmp_val);
+ msg_len -= netlbl_align(tmp_val);
+ if (ret_val != 0)
+ goto remove_return;
+ }
+
+ ret_val = 0;
+
+remove_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and dumps the domain hash table in a
+ * form suitable for use in a kernel generated LIST message. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ struct sk_buff *ans_skb;
+
+ ans_skb = netlbl_domhsh_dump(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL)
+ goto list_failure;
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_LIST,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_adddef - Handle an ADDDEF message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ADDDEF message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_adddef(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg_ptr = netlbl_netlink_payload_data(skb);
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ struct netlbl_dom_map *entry = NULL;
+ u32 tmp_val;
+ u32 rcu_locked = 0;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto adddef_failure;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto adddef_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL) {
+ ret_val = -ENOMEM;
+ goto adddef_failure;
+ }
+
+ entry->type = tmp_val;
+ switch (entry->type) {
+ case NETLBL_NLTYPE_UNLABELED:
+ break;
+ case NETLBL_NLTYPE_CIPSOV4:
+ if (msg_len < NETLBL_LEN_U32)
+ goto adddef_failure;
+ tmp_val = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ /* XXX - we should be holding a rcu_read_lock here while we
+ hold the result but since the entry will always be
+ deleted when the CIPSO DOI is deleted we are going
+ to skip the lock */
+ rcu_read_lock();
+ entry->type_def.cipsov4 = cipso_v4_doi_getdef(tmp_val);
+ if (entry->type_def.cipsov4 == NULL) {
+ rcu_read_unlock();
+ goto adddef_failure;
+ }
+ rcu_locked = 1;
+ break;
+ default:
+ ret_val = -EINVAL;
+ goto adddef_failure;
+ }
+
+ ret_val = netlbl_domhsh_add_default(entry);
+ if (rcu_locked) {
+ rcu_read_unlock();
+ rcu_locked = 0;
+ }
+ if (ret_val != 0)
+ goto adddef_failure;
+
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+
+adddef_failure:
+ kfree(entry);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_removedef - Handle a REMOVEDEF message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVEDEF message and remove the default domain
+ * mapping. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_removedef(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto removedef_return;
+
+ ret_val = netlbl_domhsh_remove_default();
+
+removedef_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_listdef - Handle a LISTDEF message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LISTDEF message and dumps the default domain
+ * mapping in a form suitable for use in a kernel generated LISTDEF message.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_listdef(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ struct sk_buff *ans_skb;
+
+ ans_skb = netlbl_domhsh_dump_default(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL)
+ goto listdef_failure;
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_LISTDEF,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto listdef_failure;
+
+ return 0;
+
+listdef_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_modules - Handle a MODULES message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated MODULES message and respond accordingly.
+ *
+ */
+static int netlbl_mgmt_modules(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ u32 mod_count;
+ size_t data_size;
+ struct sk_buff *ans_skb = NULL;
+ unsigned char *data;
+
+ /* unlabeled + cipsov4 */
+ mod_count = 2;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32 + mod_count * NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto modules_failure;
+
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_MODULES,
+ data_size);
+ if (data == NULL)
+ goto modules_failure;
+
+ netlbl_putinc_u32(&data, mod_count, NULL);
+ netlbl_putinc_u32(&data, NETLBL_NLTYPE_UNLABELED, NULL);
+ netlbl_putinc_u32(&data, NETLBL_NLTYPE_CIPSOV4, NULL);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto modules_failure;
+
+ return 0;
+
+modules_failure:
+ kfree_skb(ans_skb);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_version - Handle a VERSION message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated VERSION message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_mgmt_version(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ size_t data_size;
+ struct sk_buff *ans_skb = NULL;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto version_failure;
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_VERSION,
+ data_size);
+ if (data == NULL)
+ goto version_failure;
+
+ netlbl_put_u32(data, NETLBL_PROTO_VERSION);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto version_failure;
+
+ return 0;
+
+version_failure:
+ kfree_skb(ans_skb);
+ netlbl_netlink_send_ack(info,
+ netlbl_mgmt_gnl_family.id,
+ NLBL_MGMT_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_mgmt_genl_c_add = {
+ .cmd = NLBL_MGMT_C_ADD,
+ .flags = 0,
+ .doit = netlbl_mgmt_add,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_remove = {
+ .cmd = NLBL_MGMT_C_REMOVE,
+ .flags = 0,
+ .doit = netlbl_mgmt_remove,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_list = {
+ .cmd = NLBL_MGMT_C_LIST,
+ .flags = 0,
+ .doit = netlbl_mgmt_list,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_adddef = {
+ .cmd = NLBL_MGMT_C_ADDDEF,
+ .flags = 0,
+ .doit = netlbl_mgmt_adddef,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_removedef = {
+ .cmd = NLBL_MGMT_C_REMOVEDEF,
+ .flags = 0,
+ .doit = netlbl_mgmt_removedef,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_listdef = {
+ .cmd = NLBL_MGMT_C_LISTDEF,
+ .flags = 0,
+ .doit = netlbl_mgmt_listdef,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_modules = {
+ .cmd = NLBL_MGMT_C_MODULES,
+ .flags = 0,
+ .doit = netlbl_mgmt_modules,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_mgmt_genl_c_version = {
+ .cmd = NLBL_MGMT_C_VERSION,
+ .flags = 0,
+ .doit = netlbl_mgmt_version,
+ .dumpit = NULL,
+};
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_mgmt_genl_init - Register the NetLabel management component
+ *
+ * Description:
+ * Register the NetLabel management component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_mgmt_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_mgmt_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_add);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_remove);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_list);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_adddef);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_removedef);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_listdef);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_modules);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_version);
+ if (ret_val != 0)
+ goto genl_init_failure;
+
+ return 0;
+
+genl_init_failure:
+ netlbl_mgmt_genl_exit();
+ return ret_val;
+}
+
+/**
+ * netlbl_mgmt_genl_exit - Unregister the NetLabel management component
+ *
+ * Description:
+ * Unregister the NetLabel management component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+void netlbl_mgmt_genl_exit(void)
+{
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_version);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_modules);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_listdef);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_removedef);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_adddef);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_list);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_remove);
+ genl_unregister_ops(&netlbl_mgmt_gnl_family,
+ &netlbl_mgmt_genl_c_add);
+
+ genl_unregister_family(&netlbl_mgmt_gnl_family);
+ netlbl_mgmt_gnl_family.id = GENL_ID_GENERATE;
+}
+
Index: linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_mgmt.h
@@ -0,0 +1,250 @@
+/*
+ * NetLabel Management Support
+ *
+ * This file defines the management functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_MGMT_H
+#define _NETLABEL_MGMT_H
+
+#include <net/netlabel.h>
+
+/*
+ * The following NetLabel payloads are supported by the management interface,
+ * all of which are preceeded by the nlmsghdr struct.
+ *
+ * o ACK:
+ * Sent by the kernel in response to an applications message, applications
+ * should never send this message.
+ *
+ * +----------------------+-----------------------+
+ * | seq number (32 bits) | return code (32 bits) |
+ * +----------------------+-----------------------+
+ *
+ * seq number: the sequence number of the original message, taken from the
+ * nlmsghdr structure
+ * return code: return value, based on errno values
+ *
+ * o ADD:
+ * Sent by an application to add a domain mapping to the NetLabel system.
+ * The kernel should respond with an ACK.
+ *
+ * +-------------------+
+ * | domains (32 bits) | ...
+ * +-------------------+
+ *
+ * domains: the number of domains in the message
+ *
+ * +-------------------------+--------------------------+
+ * | domain length (32 bits) | domain string (variable) | ...
+ * +-------------------------+--------------------------+
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * domain length: the length of the "domain string" in bytes
+ * domain string: the domain string, NULL terminated
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +---------------+
+ * | doi (32 bits) |
+ * +---------------+
+ *
+ * doi: the CIPSO DOI value
+ *
+ * o REMOVE:
+ * Sent by an application to remove a domain mapping from the NetLabel
+ * system. The kernel should ACK this message.
+ *
+ * +-------------------+
+ * | domains (32 bits) | ...
+ * +-------------------+
+ *
+ * domains: the number of domains in the message
+ *
+ * +-------------------------+--------------------------+
+ * | domain length (32 bits) | domain string (variable) | ...
+ * +-------------------------+--------------------------+
+ *
+ * domain length: the length of the "domain string" in bytes
+ * domain string: the domain string, NULL terminated
+ *
+ * o LIST:
+ * This message can be sent either from an application or by the kernel in
+ * response to an application generated LIST message. When sent by an
+ * application there is no payload. The kernel should respond to a LIST
+ * message either with a LIST message on success or an ACK message on
+ * failure.
+ *
+ * +-------------------+
+ * | domains (32 bits) | ...
+ * +-------------------+
+ *
+ * domains: the number of domains in the message
+ *
+ * +-------------------------+--------------------------+
+ * | domain length (32 bits) | domain string (variable) | ...
+ * +-------------------------+--------------------------+
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * domain length: the length of the "domain string" in bytes
+ * domain string: the domain string, NULL terminated
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +----------------+---------------+
+ * | type (32 bits) | doi (32 bits) |
+ * +----------------+---------------+
+ *
+ * type: the CIPSO mapping table type (defined in the cipso_ipv4.h header
+ * as CIPSO_V4_MAP_*)
+ * doi: the CIPSO DOI value
+ *
+ * o ADDDEF:
+ * Sent by an application to set the default domain mapping for the NetLabel
+ * system. The kernel should respond with an ACK.
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +---------------+
+ * | doi (32 bits) |
+ * +---------------+
+ *
+ * doi: the CIPSO DOI value
+ *
+ * o REMOVEDEF:
+ * Sent by an application to remove the default domain mapping from the
+ * NetLabel system, there is no payload. The kernel should ACK this message.
+ *
+ * o LISTDEF:
+ * This message can be sent either from an application or by the kernel in
+ * response to an application generated LISTDEF message. When sent by an
+ * application there is no payload. The kernel should respond to a
+ * LISTDEF message either with a LISTDEF message on success or an ACK message
+ * on failure.
+ *
+ * +-------------------------+-------------- ---- --- -- -
+ * | protocol type (32 bits) | mapping data ... repeated
+ * +-------------------------+-------------- ---- --- -- -
+ *
+ * protocol type: the protocol type (defined by NETLBL_NLTYPE_*)
+ * mapping data: specific to the map type (see below)
+ *
+ * NETLBL_NLTYPE_UNLABELED
+ *
+ * No mapping data for this protocol type.
+ *
+ * NETLBL_NLTYPE_CIPSOV4
+ *
+ * +----------------+---------------+
+ * | type (32 bits) | doi (32 bits) |
+ * +----------------+---------------+
+ *
+ * type: the CIPSO mapping table type (defined in the cipso_ipv4.h header
+ * as CIPSO_V4_MAP_*)
+ * doi: the CIPSO DOI value
+ *
+ * o MODULES:
+ * Sent by an application to request a list of configured NetLabel modules
+ * in the kernel. When sent by an application there is no payload.
+ *
+ * +-------------------+
+ * | modules (32 bits) | ...
+ * +-------------------+
+ *
+ * modules: the number of modules in the message, if this is an application
+ * generated message and the value is zero then return a list of
+ * the configured modules
+ *
+ * +------------------+
+ * | module (32 bits) | ... repeated
+ * +------------------+
+ *
+ * module: the module number as defined by NETLBL_NLTYPE_*
+ *
+ * o VERSION:
+ * Sent by an application to request the NetLabel version string. When sent
+ * by an application there is no payload. This message type is also used by
+ * the kernel to respond to an VERSION request.
+ *
+ * +-------------------+
+ * | version (32 bits) |
+ * +-------------------+
+ *
+ * version: the protocol version number
+ *
+ */
+
+/* NetLabel Management commands */
+enum {
+ NLBL_MGMT_C_UNSPEC,
+ NLBL_MGMT_C_ACK,
+ NLBL_MGMT_C_ADD,
+ NLBL_MGMT_C_REMOVE,
+ NLBL_MGMT_C_LIST,
+ NLBL_MGMT_C_ADDDEF,
+ NLBL_MGMT_C_REMOVEDEF,
+ NLBL_MGMT_C_LISTDEF,
+ NLBL_MGMT_C_MODULES,
+ NLBL_MGMT_C_VERSION,
+ __NLBL_MGMT_C_MAX,
+};
+#define NLBL_MGMT_C_MAX (__NLBL_MGMT_C_MAX - 1)
+
+/* NetLabel protocol functions */
+int netlbl_mgmt_genl_init(void);
+void netlbl_mgmt_genl_exit(void);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.h
@@ -0,0 +1,99 @@
+/*
+ * NetLabel Unlabeled Support
+ *
+ * This file defines functions for dealing with unlabeled packets for the
+ * NetLabel system. The NetLabel system manages static and dynamic label
+ * mappings for network protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_UNLABELED_H
+#define _NETLABEL_UNLABELED_H
+
+#include <net/netlabel.h>
+
+/*
+ * The following NetLabel payloads are supported by the Unlabeled subsystem.
+ *
+ * o ACK:
+ * Sent by the kernel in response to an applications message, applications
+ * should never send this message.
+ *
+ * +----------------------+-----------------------+
+ * | seq number (32 bits) | return code (32 bits) |
+ * +----------------------+-----------------------+
+ *
+ * seq number: the sequence number of the original message, taken from the
+ * nlmsghdr structure
+ * return code: return value, based on errno values
+ *
+ * o ACCEPT
+ * This message is sent from an application to specify if the kernel should
+ * allow unlabled packets to pass if they do not match any of the static
+ * mappings defined in the unlabeled module.
+ *
+ * +-----------------+
+ * | allow (32 bits) |
+ * +-----------------+
+ *
+ * allow: if true (1) then allow the packets to pass, if false (0) then
+ * reject the packets
+ *
+ * o LIST
+ * This message can be sent either from an application or by the kernel in
+ * response to an application generated LIST message. When sent by an
+ * application there is no payload. The kernel should respond to a LIST
+ * message either with a LIST message on success or an ACK message on
+ * failure.
+ *
+ * +-----------------------+
+ * | accept flag (32 bits) |
+ * +-----------------------+
+ *
+ * accept flag: if true (1) then unlabeled packets are allowed to pass,
+ * if false (0) then unlabeled packets are rejected
+ *
+ */
+
+/* NetLabel Unlabeled commands */
+enum {
+ NLBL_UNLABEL_C_UNSPEC,
+ NLBL_UNLABEL_C_ACK,
+ NLBL_UNLABEL_C_ACCEPT,
+ NLBL_UNLABEL_C_LIST,
+ __NLBL_UNLABEL_C_MAX,
+};
+#define NLBL_UNLABEL_C_MAX (__NLBL_UNLABEL_C_MAX - 1)
+
+/* NetLabel protocol functions */
+int netlbl_unlabel_genl_init(void);
+void netlbl_unlabel_genl_exit(void);
+
+/* Process Unlabeled incoming network packets */
+int netlbl_unlabel_getattr(struct netlbl_lsm_secattr *secattr);
+
+/* Set the default configuration to allow Unlabeled packets */
+int netlbl_unlabel_defconf(void);
+
+#endif
Index: linux-2.6.18-rc1/net/netlabel/netlabel_user.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_user.c
@@ -0,0 +1,181 @@
+/*
+ * NetLabel NETLINK Interface
+ *
+ * This file defines the NETLINK interface for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/socket.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <asm/bug.h>
+
+#include "netlabel_mgmt.h"
+#include "netlabel_unlabeled.h"
+#include "netlabel_cipso_v4.h"
+#include "netlabel_user.h"
+
+/*
+ * NetLabel NETLINK Setup Functions
+ */
+
+/**
+ * netlbl_netlink_init - Initialize the NETLINK communication channel
+ *
+ * Description:
+ * Call out to the NetLabel components so they can register their families and
+ * commands with the Generic NETLINK mechanism. Returns zero on success and
+ * non-zero on failure.
+ *
+ */
+int netlbl_netlink_init(void)
+{
+ int ret_val;
+
+ ret_val = netlbl_mgmt_genl_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_cipsov4_genl_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ ret_val = netlbl_unlabel_genl_init();
+ if (ret_val != 0)
+ goto init_failure;
+
+ return 0;
+
+init_failure:
+ netlbl_unlabel_genl_exit();
+ return ret_val;
+}
+
+/**
+ * netlbl_exit - Shutdown the NETLINK communication channel
+ *
+ * Description:
+ * Call out to the NetLabel components so they can deregister themselves from
+ * the Generic NETLINK mechanism.
+ *
+ */
+void netlbl_netlink_exit(void)
+{
+ netlbl_unlabel_genl_exit();
+ netlbl_cipsov4_genl_exit();
+ netlbl_mgmt_genl_exit();
+}
+
+/*
+ * NetLabel Common Protocol Functions
+ */
+
+/**
+ * netlbl_netlink_send_ack - Send an ACK message
+ * @info: the generic NETLINK information
+ * @genl_family: the generic NETLINK family ID value
+ * @ack_cmd: the generic NETLINK family ACK command value
+ * @ret_code: return code to use
+ *
+ * Description:
+ * This function sends an ACK message to the sender of the NETLINK message
+ * specified by @info.
+ *
+ */
+void netlbl_netlink_send_ack(const struct genl_info *info,
+ const u32 genl_family,
+ const u8 ack_cmd,
+ const u32 ret_code)
+{
+ size_t data_size;
+ struct sk_buff *skb;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + 2 * NETLBL_LEN_U32;
+ skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (skb == NULL)
+ return;
+
+ data = netlbl_netlink_hdr_put(skb,
+ info->snd_pid,
+ 0,
+ 0,
+ genl_family,
+ ack_cmd,
+ data_size);
+ if (data == NULL)
+ goto send_ack_failure;
+
+ netlbl_putinc_u32(&data, info->snd_seq, NULL);
+ netlbl_putinc_u32(&data, ret_code, NULL);
+
+ netlbl_netlink_snd(skb, info->snd_pid);
+ return;
+
+send_ack_failure:
+ kfree_skb(skb);
+}
+
+/*
+ * NETLINK I/O Functions
+ */
+
+/**
+ * netlbl_netlink_snd - Send a NetLabel message
+ * @skb: NetLabel message
+ * @pid: destination PID
+ *
+ * Description:
+ * Sends a unicast NetLabel message over the NETLINK socket.
+ *
+ */
+int netlbl_netlink_snd(struct sk_buff *skb, const u32 pid)
+{
+ return genlmsg_unicast(skb, pid);
+}
+
+/**
+ * netlbl_netlink_snd - Send a NetLabel message
+ * @skb: NetLabel message
+ * @pid: sending PID
+ * @group: multicast group id
+ *
+ * Description:
+ * Sends a multicast NetLabel message over the NETLINK socket to all members
+ * of @group except @pid.
+ *
+ */
+int netlbl_netlink_snd_multicast(struct sk_buff *skb,
+ const u32 pid,
+ const u32 group)
+{
+ return genlmsg_multicast(skb, pid, group);
+}
Index: linux-2.6.18-rc1/net/netlabel/netlabel_user.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_user.h
@@ -0,0 +1,210 @@
+/*
+ * NetLabel NETLINK Interface
+ *
+ * This file defines the NETLINK interface for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_USER_H
+#define _NETLABEL_USER_H
+
+#include <linux/skbuff.h>
+#include <linux/capability.h>
+#include <linux/genetlink.h>
+#include <net/netlabel.h>
+#include <net/genetlink.h>
+
+/* NetLabel NETLINK helper functions */
+
+/**
+ * netlbl_netlink_cap_check - Check the NETLINK msg capabilities
+ * @skb: the NETLINK buffer
+ * @req_cap: the required capability
+ *
+ * Description:
+ * Check the NETLINK buffer's capabilities against the required capabilities.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static inline int netlbl_netlink_cap_check(const struct sk_buff *skb,
+ const kernel_cap_t req_cap)
+{
+ if (cap_raised(NETLINK_CB(skb).eff_cap, req_cap))
+ return 0;
+ return -EPERM;
+}
+
+/**
+ * netlbl_netlink_hdr_put - Write the NETLINK buffers into a buffer
+ * @skb: the buffer
+ * @pid: the PID of the receipient
+ * @seq: the sequence number
+ * @flags: NETLINK message flags
+ * @type: the generic NETLINK message family type
+ * @cmd: command
+ * @len: length
+ *
+ * Description:
+ * Write both a NETLINK nlmsghdr structure and a Generic NETLINK genlmsghdr
+ * struct to a buffer. Returns a pointer to the start of the payload buffer
+ * on success or NULL on failure.
+ *
+ */
+static inline void __netlbl_netlink_hdr(unsigned char *buffer,
+ const u32 pid,
+ const u32 seq,
+ const int flags,
+ const int type,
+ const u8 cmd,
+ const int len)
+{
+ struct nlmsghdr *nlh;
+ struct genlmsghdr *hdr;
+
+ nlh = (struct nlmsghdr *)buffer;
+ nlh->nlmsg_type = type;
+ nlh->nlmsg_len = len;
+ nlh->nlmsg_flags = flags;
+ nlh->nlmsg_pid = pid;
+ nlh->nlmsg_seq = seq;
+
+ hdr = nlmsg_data(nlh);
+ hdr->cmd = cmd;
+ hdr->version = NETLBL_PROTO_VERSION;
+ hdr->reserved = 0;
+}
+
+/**
+ * netlbl_netlink_hdr_put - Write the NETLINK buffers into a sk_buff
+ * @skb: the packet
+ * @pid: the PID of the receipient
+ * @seq: the sequence number
+ * @flags: NETLINK message flags
+ * @type: the generic NETLINK message family type
+ * @cmd: command
+ * @len: length
+ *
+ * Description:
+ * Write both a NETLINK nlmsghdr structure and a Generic NETLINK genlmsghdr
+ * struct to the packet. Returns a pointer to the start of the payload buffer
+ * on success or NULL on failure.
+ *
+ */
+static inline unsigned char *netlbl_netlink_hdr_put(struct sk_buff *skb,
+ const u32 pid,
+ const u32 seq,
+ const int flags,
+ const int type,
+ const u8 cmd,
+ const int len)
+{
+ struct nlmsghdr *nlh;
+ struct genlmsghdr *hdr;
+
+ nlh = nlmsg_put(skb, pid, seq, type, len, flags);
+ if (nlh == NULL)
+ return NULL;
+
+ hdr = nlmsg_data(nlh);
+ hdr->cmd = cmd;
+ hdr->version = NETLBL_PROTO_VERSION;
+ hdr->reserved = 0;
+
+ return (unsigned char *)hdr + GENL_HDRLEN;
+}
+
+/**
+ * netlbl_netlink_hdr_push - Write the NETLINK buffers into a sk_buff
+ * @skb: the packet
+ * @pid: the PID of the receipient
+ * @seq: the sequence number
+ * @flags: NETLINK message flags
+ * @type: the generic NETLINK message family type
+ * @cmd: command
+ * @len: length
+ *
+ * Description:
+ * Write both a NETLINK nlmsghdr structure and a Generic NETLINK genlmsghdr
+ * struct to the packet.
+ *
+ */
+static inline void netlbl_netlink_hdr_push(struct sk_buff *skb,
+ const u32 pid,
+ const u32 seq,
+ const int flags,
+ const int type,
+ const u8 cmd,
+ const int len)
+
+{
+ unsigned char *buf = skb_push(skb, NLMSG_SPACE(GENL_HDRLEN));
+ __netlbl_netlink_hdr(buf, pid, seq, flags, type, cmd, len);
+}
+
+/**
+ * netlbl_netlink_payload_len - Return the length of the payload
+ * @skb: the NETLINK buffer
+ *
+ * Description:
+ * This function returns the length of the NetLabel payload.
+ *
+ */
+static inline u32 netlbl_netlink_payload_len(const struct sk_buff *skb)
+{
+ return nlmsg_len((struct nlmsghdr *)skb->data) - GENL_HDRLEN;
+}
+
+/**
+ * netlbl_netlink_payload_data - Returns a pointer to the start of the payload
+ * @skb: the NETLINK buffer
+ *
+ * Description:
+ * This function returns a pointer to the start of the NetLabel payload.
+ *
+ */
+static inline unsigned char *netlbl_netlink_payload_data(
+ const struct sk_buff *skb)
+{
+ return nlmsg_data((struct nlmsghdr *)skb->data) + GENL_HDRLEN;
+}
+
+/* NetLabel common protocol functions */
+
+void netlbl_netlink_send_ack(const struct genl_info *info,
+ const u32 genl_family,
+ const u8 ack_cmd,
+ const u32 ret_code);
+
+/* NetLabel NETLINK I/O functions */
+
+int netlbl_netlink_init(void);
+void netlbl_netlink_exit(void);
+int netlbl_netlink_snd(struct sk_buff *skb, const u32 pid);
+int netlbl_netlink_snd_multicast(struct sk_buff *skb,
+ const u32 pid,
+ const u32 group);
+
+#endif
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 5/7] NetLabel: CIPSOv4 and Unlabeled packet integration
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-core_components --]
[-- Type: text/plain, Size: 23849 bytes --]
Add CIPSO/IPv4 and unlabeled packet management to the NetLabel subsystem. The
CIPSO/IPv4 changes allow the configuration of CIPSO/IPv4 within the overall
NetLabel framework. The unlabeled packet changes allows NetLabel to pass
unlabeled packets without error.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/netlabel/netlabel_cipso_v4.c | 582 ++++++++++++++++++++++++++++++++++++++
net/netlabel/netlabel_unlabeled.c | 272 +++++++++++++++++
2 files changed, 854 insertions(+)
Index: linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.c
@@ -0,0 +1,582 @@
+/*
+ * NetLabel CIPSO/IPv4 Support
+ *
+ * This file defines the CIPSO/IPv4 functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+
+#include "netlabel_user.h"
+#include "netlabel_cipso_v4.h"
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_cipsov4_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_CIPSOV4_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * netlbl_cipsov4_doi_free - Frees a CIPSO V4 DOI definition
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to the DOI definition can be released
+ * safely.
+ *
+ */
+static void netlbl_cipsov4_doi_free(struct rcu_head *entry)
+{
+ struct cipso_v4_doi *ptr;
+
+ ptr = container_of(entry, struct cipso_v4_doi, rcu);
+ switch (ptr->type) {
+ case CIPSO_V4_MAP_STD:
+ kfree(ptr->map.std->lvl.cipso);
+ kfree(ptr->map.std->lvl.local);
+ kfree(ptr->map.std->cat.cipso);
+ kfree(ptr->map.std->cat.local);
+ break;
+ }
+ kfree(ptr);
+}
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_cipsov4_add_std - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_STD DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_std(const u32 doi,
+ const unsigned char *msg,
+ const size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ u32 num_lvls;
+ u32 num_cats;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+ u32 tmp_val_a;
+ u32 tmp_val_b;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_std_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_std_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std = kzalloc(sizeof(*doi_def->map.std), GFP_KERNEL);
+ if (doi_def->map.std == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_STD;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_std_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_std_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ if (msg_len < 6 * NETLBL_LEN_U32)
+ goto add_std_failure;
+
+ num_lvls = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_lvls == 0)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.local_size > CIPSO_V4_MAX_LOC_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local = kcalloc(doi_def->map.std->lvl.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->lvl.cipso_size = netlbl_getinc_u8(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.cipso_size > CIPSO_V4_MAX_REM_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.cipso = kcalloc(doi_def->map.std->lvl.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ num_cats = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ doi_def->map.std->cat.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.local_size > CIPSO_V4_MAX_LOC_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.local = kcalloc(doi_def->map.std->cat.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->cat.cipso_size = netlbl_getinc_u16(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.cipso_size > CIPSO_V4_MAX_REM_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.cipso = kcalloc(doi_def->map.std->cat.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ if (msg_len <
+ num_lvls * (NETLBL_LEN_U32 + NETLBL_LEN_U8) +
+ num_cats * (NETLBL_LEN_U32 + NETLBL_LEN_U16))
+ goto add_std_failure;
+
+ for (iter = 0; iter < doi_def->map.std->lvl.cipso_size; iter++)
+ doi_def->map.std->lvl.cipso[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->lvl.local_size; iter++)
+ doi_def->map.std->lvl.local[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->cat.cipso_size; iter++)
+ doi_def->map.std->cat.cipso[iter] = CIPSO_V4_INV_CAT;
+ for (iter = 0; iter < doi_def->map.std->cat.local_size; iter++)
+ doi_def->map.std->cat.local[iter] = CIPSO_V4_INV_CAT;
+
+ for (iter = 0; iter < num_lvls; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u8(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->lvl.local_size ||
+ tmp_val_b >= doi_def->map.std->lvl.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->lvl.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->lvl.local[tmp_val_a] = tmp_val_b;
+ }
+
+ for (iter = 0; iter < num_cats; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u16(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->cat.local_size ||
+ tmp_val_b >= doi_def->map.std->cat.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->cat.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->cat.local[tmp_val_a] = tmp_val_b;
+ }
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_std_failure;
+ return 0;
+
+add_std_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add_pass - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_PASS DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_pass(const u32 doi,
+ const unsigned char *msg,
+ const size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_pass_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_pass_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_pass_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_PASS;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_pass_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_pass_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_pass_failure;
+ return 0;
+
+add_pass_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add - Handle an ADD message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Create a new DOI definition based on the given ADD message and add it to the
+ * CIPSO V4 engine. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_add(struct sk_buff *skb, struct genl_info *info)
+
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ u32 map_type;
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto add_return;
+
+ if (msg_len < 2 * NETLBL_LEN_U32)
+ goto add_return;
+
+ doi = netlbl_getinc_u32(&msg, &msg_len);
+ map_type = netlbl_getinc_u32(&msg, &msg_len);
+ switch (map_type) {
+ case CIPSO_V4_MAP_STD:
+ ret_val = netlbl_cipsov4_add_std(doi, msg, msg_len);
+ break;
+ case CIPSO_V4_MAP_PASS:
+ ret_val = netlbl_cipsov4_add_pass(doi, msg, msg_len);
+ break;
+ }
+
+add_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ struct sk_buff *ans_skb;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32)
+ goto list_failure;
+
+ doi = netlbl_get_u32(msg);
+ ans_skb = cipso_v4_doi_dump(doi, NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto list_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LIST,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_listall - Handle a LISTALL message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LISTALL message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_listall(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ struct sk_buff *ans_skb;
+
+ ans_skb = cipso_v4_doi_dump_all(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto listall_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LISTALL,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto listall_failure;
+
+ return 0;
+
+listall_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_remove - Handle a REMOVE message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVE message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_remove(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto remove_return;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32) {
+ ret_val = -EINVAL;
+ goto remove_return;
+ }
+
+ doi = netlbl_get_u32(msg);
+ ret_val = cipso_v4_doi_remove(doi, netlbl_cipsov4_doi_free);
+
+remove_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_cipsov4_genl_c_add = {
+ .cmd = NLBL_CIPSOV4_C_ADD,
+ .flags = 0,
+ .doit = netlbl_cipsov4_add,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_remove = {
+ .cmd = NLBL_CIPSOV4_C_REMOVE,
+ .flags = 0,
+ .doit = netlbl_cipsov4_remove,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_list = {
+ .cmd = NLBL_CIPSOV4_C_LIST,
+ .flags = 0,
+ .doit = netlbl_cipsov4_list,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_listall = {
+ .cmd = NLBL_CIPSOV4_C_LISTALL,
+ .flags = 0,
+ .doit = netlbl_cipsov4_listall,
+ .dumpit = NULL,
+};
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_cipsov4_genl_init - Register the CIPSOv4 NetLabel component
+ *
+ * Description:
+ * Register the CIPSOv4 packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_cipsov4_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_cipsov4_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_add);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_remove);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_list);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_listall);
+ if (ret_val != 0)
+ goto genl_init_failure;
+
+ return 0;
+
+genl_init_failure:
+ netlbl_cipsov4_genl_exit();
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_genl_exit - Unregister the CIPSOv4 NetLabel component
+ *
+ * Description:
+ * Unregister the CIPSOv4 packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+void netlbl_cipsov4_genl_exit(void)
+{
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_listall);
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_list);
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_remove);
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_add);
+
+ genl_unregister_family(&netlbl_cipsov4_gnl_family);
+ netlbl_cipsov4_gnl_family.id = GENL_ID_GENERATE;
+}
+
Index: linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.c
@@ -0,0 +1,272 @@
+/*
+ * NetLabel Unlabeled Support
+ *
+ * This file defines functions for dealing with unlabeled packets for the
+ * NetLabel system. The NetLabel system manages static and dynamic label
+ * mappings for network protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+#include <net/netlabel.h>
+#include <asm/bug.h>
+
+#include "netlabel_user.h"
+#include "netlabel_domainhash.h"
+#include "netlabel_unlabeled.h"
+
+/* Accept unlabeled packets flag */
+static atomic_t netlabel_unlabel_accept_flg = ATOMIC_INIT(0);
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_unlabel_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_UNLABELED_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_unlabel_accept - Handle an ACCEPT message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ACCEPT message and set the accept flag accordingly.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_accept(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ u32 value;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ return ret_val;
+
+ if (netlbl_netlink_payload_len(skb) == NETLBL_LEN_U32) {
+ value = netlbl_get_u32(msg);
+ if (value == 1 || value == 0) {
+ atomic_set(&netlabel_unlabel_accept_flg, value);
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+ }
+ }
+
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ EINVAL);
+ return -EINVAL;
+}
+
+/**
+ * netlbl_unlabel_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond with the current status.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ size_t data_size;
+ struct sk_buff *ans_skb;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto list_failure;
+
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_LIST,
+ data_size);
+ if (data == NULL)
+ goto list_failure;
+
+ netlbl_put_u32(data, atomic_read(&netlabel_unlabel_accept_flg));
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_unlabel_genl_c_accept = {
+ .cmd = NLBL_UNLABEL_C_ACCEPT,
+ .flags = 0,
+ .doit = netlbl_unlabel_accept,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_unlabel_genl_c_list = {
+ .cmd = NLBL_UNLABEL_C_LIST,
+ .flags = 0,
+ .doit = netlbl_unlabel_list,
+ .dumpit = NULL,
+};
+
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_unlabel_genl_init - Register the Unlabeled NetLabel component
+ *
+ * Description:
+ * Register the unlabeled packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_unlabel_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_unlabel_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_accept);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_list);
+ if (ret_val != 0)
+ return ret_val;
+
+ return 0;
+}
+
+/**
+ * netlbl_unlabel_genl_exit - Unregister the Unlabeled NetLabel component
+ *
+ * Description:
+ * Unregister the unlabeled packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+void netlbl_unlabel_genl_exit(void)
+{
+ genl_unregister_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_list);
+ genl_unregister_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_accept);
+
+ genl_unregister_family(&netlbl_unlabel_gnl_family);
+ netlbl_unlabel_gnl_family.id = GENL_ID_GENERATE;
+}
+
+/*
+ * NetLabel KAPI Hooks
+ */
+
+/**
+ * netlbl_unlabel_getattr - Get the security attributes for an unlabled packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Determine the security attributes, if any, for an unlabled packet and return
+ * them in @secattr. Returns zero on success and negative values on failure.
+ *
+ */
+int netlbl_unlabel_getattr(struct netlbl_lsm_secattr *secattr)
+{
+ if (atomic_read(&netlabel_unlabel_accept_flg) == 1) {
+ memset(secattr, 0, sizeof(*secattr));
+ return 0;
+ }
+
+ return -ENOMSG;
+}
+
+/**
+ * netlbl_unlabel_defconf - Set the default config to allow unlabeled packets
+ *
+ * Description:
+ * Set the default NetLabel configuration to allow incoming unlabeled packets
+ * and to send unlabeled network traffic by default.
+ *
+ */
+int netlbl_unlabel_defconf(void)
+{
+ int ret_val;
+ struct netlbl_dom_map *entry;
+
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL)
+ return -ENOMEM;
+ entry->type = NETLBL_NLTYPE_UNLABELED;
+ ret_val = netlbl_domhsh_add_default(entry);
+ if (ret_val != 0)
+ return ret_val;
+
+ atomic_set(&netlabel_unlabel_accept_flg, 1);
+
+ return 0;
+}
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 5/7] NetLabel: CIPSOv4 and Unlabeled packet integration
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Add CIPSO/IPv4 and unlabeled packet management to the NetLabel subsystem. The
CIPSO/IPv4 changes allow the configuration of CIPSO/IPv4 within the overall
NetLabel framework. The unlabeled packet changes allows NetLabel to pass
unlabeled packets without error.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/netlabel/netlabel_cipso_v4.c | 582 ++++++++++++++++++++++++++++++++++++++
net/netlabel/netlabel_unlabeled.c | 272 +++++++++++++++++
2 files changed, 854 insertions(+)
Index: linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_cipso_v4.c
@@ -0,0 +1,582 @@
+/*
+ * NetLabel CIPSO/IPv4 Support
+ *
+ * This file defines the CIPSO/IPv4 functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+
+#include "netlabel_user.h"
+#include "netlabel_cipso_v4.h"
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_cipsov4_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_CIPSOV4_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * netlbl_cipsov4_doi_free - Frees a CIPSO V4 DOI definition
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to the DOI definition can be released
+ * safely.
+ *
+ */
+static void netlbl_cipsov4_doi_free(struct rcu_head *entry)
+{
+ struct cipso_v4_doi *ptr;
+
+ ptr = container_of(entry, struct cipso_v4_doi, rcu);
+ switch (ptr->type) {
+ case CIPSO_V4_MAP_STD:
+ kfree(ptr->map.std->lvl.cipso);
+ kfree(ptr->map.std->lvl.local);
+ kfree(ptr->map.std->cat.cipso);
+ kfree(ptr->map.std->cat.local);
+ break;
+ }
+ kfree(ptr);
+}
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_cipsov4_add_std - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_STD DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_std(const u32 doi,
+ const unsigned char *msg,
+ const size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ u32 num_lvls;
+ u32 num_cats;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+ u32 tmp_val_a;
+ u32 tmp_val_b;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_std_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_std_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std = kzalloc(sizeof(*doi_def->map.std), GFP_KERNEL);
+ if (doi_def->map.std == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_STD;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_std_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_std_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ if (msg_len < 6 * NETLBL_LEN_U32)
+ goto add_std_failure;
+
+ num_lvls = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_lvls == 0)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.local_size > CIPSO_V4_MAX_LOC_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local = kcalloc(doi_def->map.std->lvl.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->lvl.cipso_size = netlbl_getinc_u8(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.cipso_size > CIPSO_V4_MAX_REM_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.cipso = kcalloc(doi_def->map.std->lvl.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ num_cats = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ doi_def->map.std->cat.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.local_size > CIPSO_V4_MAX_LOC_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.local = kcalloc(doi_def->map.std->cat.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->cat.cipso_size = netlbl_getinc_u16(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.cipso_size > CIPSO_V4_MAX_REM_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.cipso = kcalloc(doi_def->map.std->cat.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ if (msg_len <
+ num_lvls * (NETLBL_LEN_U32 + NETLBL_LEN_U8) +
+ num_cats * (NETLBL_LEN_U32 + NETLBL_LEN_U16))
+ goto add_std_failure;
+
+ for (iter = 0; iter < doi_def->map.std->lvl.cipso_size; iter++)
+ doi_def->map.std->lvl.cipso[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->lvl.local_size; iter++)
+ doi_def->map.std->lvl.local[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->cat.cipso_size; iter++)
+ doi_def->map.std->cat.cipso[iter] = CIPSO_V4_INV_CAT;
+ for (iter = 0; iter < doi_def->map.std->cat.local_size; iter++)
+ doi_def->map.std->cat.local[iter] = CIPSO_V4_INV_CAT;
+
+ for (iter = 0; iter < num_lvls; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u8(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->lvl.local_size ||
+ tmp_val_b >= doi_def->map.std->lvl.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->lvl.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->lvl.local[tmp_val_a] = tmp_val_b;
+ }
+
+ for (iter = 0; iter < num_cats; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u16(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->cat.local_size ||
+ tmp_val_b >= doi_def->map.std->cat.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->cat.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->cat.local[tmp_val_a] = tmp_val_b;
+ }
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_std_failure;
+ return 0;
+
+add_std_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add_pass - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_PASS DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_pass(const u32 doi,
+ const unsigned char *msg,
+ const size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_pass_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_pass_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_pass_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_PASS;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_pass_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_pass_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_pass_failure;
+ return 0;
+
+add_pass_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add - Handle an ADD message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Create a new DOI definition based on the given ADD message and add it to the
+ * CIPSO V4 engine. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_add(struct sk_buff *skb, struct genl_info *info)
+
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ u32 map_type;
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto add_return;
+
+ if (msg_len < 2 * NETLBL_LEN_U32)
+ goto add_return;
+
+ doi = netlbl_getinc_u32(&msg, &msg_len);
+ map_type = netlbl_getinc_u32(&msg, &msg_len);
+ switch (map_type) {
+ case CIPSO_V4_MAP_STD:
+ ret_val = netlbl_cipsov4_add_std(doi, msg, msg_len);
+ break;
+ case CIPSO_V4_MAP_PASS:
+ ret_val = netlbl_cipsov4_add_pass(doi, msg, msg_len);
+ break;
+ }
+
+add_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ struct sk_buff *ans_skb;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32)
+ goto list_failure;
+
+ doi = netlbl_get_u32(msg);
+ ans_skb = cipso_v4_doi_dump(doi, NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto list_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LIST,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_listall - Handle a LISTALL message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LISTALL message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_listall(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ struct sk_buff *ans_skb;
+
+ ans_skb = cipso_v4_doi_dump_all(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto listall_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LISTALL,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto listall_failure;
+
+ return 0;
+
+listall_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_remove - Handle a REMOVE message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVE message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_remove(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto remove_return;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32) {
+ ret_val = -EINVAL;
+ goto remove_return;
+ }
+
+ doi = netlbl_get_u32(msg);
+ ret_val = cipso_v4_doi_remove(doi, netlbl_cipsov4_doi_free);
+
+remove_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_cipsov4_genl_c_add = {
+ .cmd = NLBL_CIPSOV4_C_ADD,
+ .flags = 0,
+ .doit = netlbl_cipsov4_add,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_remove = {
+ .cmd = NLBL_CIPSOV4_C_REMOVE,
+ .flags = 0,
+ .doit = netlbl_cipsov4_remove,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_list = {
+ .cmd = NLBL_CIPSOV4_C_LIST,
+ .flags = 0,
+ .doit = netlbl_cipsov4_list,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_listall = {
+ .cmd = NLBL_CIPSOV4_C_LISTALL,
+ .flags = 0,
+ .doit = netlbl_cipsov4_listall,
+ .dumpit = NULL,
+};
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_cipsov4_genl_init - Register the CIPSOv4 NetLabel component
+ *
+ * Description:
+ * Register the CIPSOv4 packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_cipsov4_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_cipsov4_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_add);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_remove);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_list);
+ if (ret_val != 0)
+ goto genl_init_failure;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_listall);
+ if (ret_val != 0)
+ goto genl_init_failure;
+
+ return 0;
+
+genl_init_failure:
+ netlbl_cipsov4_genl_exit();
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_genl_exit - Unregister the CIPSOv4 NetLabel component
+ *
+ * Description:
+ * Unregister the CIPSOv4 packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+void netlbl_cipsov4_genl_exit(void)
+{
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_listall);
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_list);
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_remove);
+ genl_unregister_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_add);
+
+ genl_unregister_family(&netlbl_cipsov4_gnl_family);
+ netlbl_cipsov4_gnl_family.id = GENL_ID_GENERATE;
+}
+
Index: linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/net/netlabel/netlabel_unlabeled.c
@@ -0,0 +1,272 @@
+/*
+ * NetLabel Unlabeled Support
+ *
+ * This file defines functions for dealing with unlabeled packets for the
+ * NetLabel system. The NetLabel system manages static and dynamic label
+ * mappings for network protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+#include <net/netlabel.h>
+#include <asm/bug.h>
+
+#include "netlabel_user.h"
+#include "netlabel_domainhash.h"
+#include "netlabel_unlabeled.h"
+
+/* Accept unlabeled packets flag */
+static atomic_t netlabel_unlabel_accept_flg = ATOMIC_INIT(0);
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_unlabel_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_UNLABELED_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_unlabel_accept - Handle an ACCEPT message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ACCEPT message and set the accept flag accordingly.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_accept(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ u32 value;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ return ret_val;
+
+ if (netlbl_netlink_payload_len(skb) == NETLBL_LEN_U32) {
+ value = netlbl_get_u32(msg);
+ if (value == 1 || value == 0) {
+ atomic_set(&netlabel_unlabel_accept_flg, value);
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+ }
+ }
+
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ EINVAL);
+ return -EINVAL;
+}
+
+/**
+ * netlbl_unlabel_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond with the current status.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ size_t data_size;
+ struct sk_buff *ans_skb;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto list_failure;
+
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_LIST,
+ data_size);
+ if (data == NULL)
+ goto list_failure;
+
+ netlbl_put_u32(data, atomic_read(&netlabel_unlabel_accept_flg));
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_unlabel_genl_c_accept = {
+ .cmd = NLBL_UNLABEL_C_ACCEPT,
+ .flags = 0,
+ .doit = netlbl_unlabel_accept,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_unlabel_genl_c_list = {
+ .cmd = NLBL_UNLABEL_C_LIST,
+ .flags = 0,
+ .doit = netlbl_unlabel_list,
+ .dumpit = NULL,
+};
+
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_unlabel_genl_init - Register the Unlabeled NetLabel component
+ *
+ * Description:
+ * Register the unlabeled packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_unlabel_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_unlabel_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_accept);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_list);
+ if (ret_val != 0)
+ return ret_val;
+
+ return 0;
+}
+
+/**
+ * netlbl_unlabel_genl_exit - Unregister the Unlabeled NetLabel component
+ *
+ * Description:
+ * Unregister the unlabeled packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+void netlbl_unlabel_genl_exit(void)
+{
+ genl_unregister_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_list);
+ genl_unregister_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_accept);
+
+ genl_unregister_family(&netlbl_unlabel_gnl_family);
+ netlbl_unlabel_gnl_family.id = GENL_ID_GENERATE;
+}
+
+/*
+ * NetLabel KAPI Hooks
+ */
+
+/**
+ * netlbl_unlabel_getattr - Get the security attributes for an unlabled packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Determine the security attributes, if any, for an unlabled packet and return
+ * them in @secattr. Returns zero on success and negative values on failure.
+ *
+ */
+int netlbl_unlabel_getattr(struct netlbl_lsm_secattr *secattr)
+{
+ if (atomic_read(&netlabel_unlabel_accept_flg) == 1) {
+ memset(secattr, 0, sizeof(*secattr));
+ return 0;
+ }
+
+ return -ENOMSG;
+}
+
+/**
+ * netlbl_unlabel_defconf - Set the default config to allow unlabeled packets
+ *
+ * Description:
+ * Set the default NetLabel configuration to allow incoming unlabeled packets
+ * and to send unlabeled network traffic by default.
+ *
+ */
+int netlbl_unlabel_defconf(void)
+{
+ int ret_val;
+ struct netlbl_dom_map *entry;
+
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL)
+ return -ENOMEM;
+ entry->type = NETLBL_NLTYPE_UNLABELED;
+ ret_val = netlbl_domhsh_add_default(entry);
+ if (ret_val != 0)
+ return ret_val;
+
+ atomic_set(&netlabel_unlabel_accept_flg, 1);
+
+ return 0;
+}
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 6/7] NetLabel: SELinux support
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-selinux-2.6.18 --]
[-- Type: text/plain, Size: 41460 bytes --]
Add NetLabel support to the SELinux LSM and modify the socket_post_create() LSM
hook to return an error code. The most significant part of this patch is the
addition of NetLabel hooks into the following SELinux LSM hooks:
* selinux_file_permission()
* selinux_socket_sendmsg()
* selinux_socket_post_create()
* selinux_socket_accept()
* selinux_socket_post_accept() [NEW]
* selinux_socket_sock_rcv_skb()
* selinux_socket_getpeersec_stream()
* selinux_socket_getpeersec_dgram()
The basic reasoning behind this patch is that outgoing packets are "NetLabel'd"
by labeling their socket and the NetLabel security attributes are checked via
the additional hook in selinux_socket_sock_rcv_skb(). NetLabel itself is only
a labeling mechanism, similar to filesystem extended attributes, it is up to
the SELinux enforcement mechanism to perform the actual access checks.
In addition to the changes outlined above this patch also includes some changes
to the extended bitmap (ebitmap) and multi-level security (mls) code to import
and export SELinux TE/MLS attributes into and out of NetLabel.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
include/linux/security.h | 25 -
net/socket.c | 17
security/dummy.c | 6
security/selinux/hooks.c | 53 ++
security/selinux/include/objsec.h | 13
security/selinux/include/selinux_netlabel.h | 100 +++++
security/selinux/ss/ebitmap.c | 152 +++++++
security/selinux/ss/ebitmap.h | 6
security/selinux/ss/mls.c | 152 +++++++
security/selinux/ss/mls.h | 21 +
security/selinux/ss/services.c | 555 ++++++++++++++++++++++++++++
11 files changed, 1070 insertions(+), 30 deletions(-)
Index: linux-2.6.18-rc1/include/linux/security.h
===================================================================
--- linux-2.6.18-rc1.orig/include/linux/security.h
+++ linux-2.6.18-rc1/include/linux/security.h
@@ -1296,8 +1296,8 @@ struct security_operations {
int (*unix_may_send) (struct socket * sock, struct socket * other);
int (*socket_create) (int family, int type, int protocol, int kern);
- void (*socket_post_create) (struct socket * sock, int family,
- int type, int protocol, int kern);
+ int (*socket_post_create) (struct socket * sock, int family,
+ int type, int protocol, int kern);
int (*socket_bind) (struct socket * sock,
struct sockaddr * address, int addrlen);
int (*socket_connect) (struct socket * sock,
@@ -2748,13 +2748,13 @@ static inline int security_socket_create
return security_ops->socket_create(family, type, protocol, kern);
}
-static inline void security_socket_post_create(struct socket * sock,
- int family,
- int type,
- int protocol, int kern)
+static inline int security_socket_post_create(struct socket * sock,
+ int family,
+ int type,
+ int protocol, int kern)
{
- security_ops->socket_post_create(sock, family, type,
- protocol, kern);
+ return security_ops->socket_post_create(sock, family, type,
+ protocol, kern);
}
static inline int security_socket_bind(struct socket * sock,
@@ -2880,11 +2880,12 @@ static inline int security_socket_create
return 0;
}
-static inline void security_socket_post_create(struct socket * sock,
- int family,
- int type,
- int protocol, int kern)
+static inline int security_socket_post_create(struct socket * sock,
+ int family,
+ int type,
+ int protocol, int kern)
{
+ return 0;
}
static inline int security_socket_bind(struct socket * sock,
Index: linux-2.6.18-rc1/net/socket.c
===================================================================
--- linux-2.6.18-rc1.orig/net/socket.c
+++ linux-2.6.18-rc1/net/socket.c
@@ -973,11 +973,18 @@ int sock_create_lite(int family, int typ
goto out;
}
- security_socket_post_create(sock, family, type, protocol, 1);
sock->type = type;
+ err = security_socket_post_create(sock, family, type, protocol, 1);
+ if (err)
+ goto out_release;
+
out:
*res = sock;
return err;
+out_release:
+ sock_release(sock);
+ sock = NULL;
+ goto out;
}
/* No kernel lock held - perfect */
@@ -1213,7 +1220,9 @@ static int __sock_create(int family, int
*/
module_put(net_families[family]->owner);
*res = sock;
- security_socket_post_create(sock, family, type, protocol, kern);
+ err = security_socket_post_create(sock, family, type, protocol, kern);
+ if (err)
+ goto out_release;
out:
net_family_read_unlock();
@@ -1443,13 +1452,13 @@ asmlinkage long sys_accept(int fd, struc
goto out_fd;
}
+ security_socket_post_accept(sock, newsock);
+
/* File flags are not inherited via accept() unlike another OSes. */
fd_install(newfd, newfile);
err = newfd;
- security_socket_post_accept(sock, newsock);
-
out_put:
fput_light(sock->file, fput_needed);
out:
Index: linux-2.6.18-rc1/security/dummy.c
===================================================================
--- linux-2.6.18-rc1.orig/security/dummy.c
+++ linux-2.6.18-rc1/security/dummy.c
@@ -709,10 +709,10 @@ static int dummy_socket_create (int fami
return 0;
}
-static void dummy_socket_post_create (struct socket *sock, int family, int type,
- int protocol, int kern)
+static int dummy_socket_post_create (struct socket *sock, int family, int type,
+ int protocol, int kern)
{
- return;
+ return 0;
}
static int dummy_socket_bind (struct socket *sock, struct sockaddr *address,
Index: linux-2.6.18-rc1/security/selinux/hooks.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/hooks.c
+++ linux-2.6.18-rc1/security/selinux/hooks.c
@@ -12,6 +12,8 @@
* Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
* Copyright (C) 2004-2005 Trusted Computer Solutions, Inc.
* <dgoeddel@trustedcs.com>
+ * Copyright (C) 2006 Hewlett-Packard Development Company, L.P.
+ * Paul Moore, <paul.moore@hp.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2,
@@ -74,6 +76,7 @@
#include "objsec.h"
#include "netif.h"
#include "xfrm.h"
+#include "selinux_netlabel.h"
#define XATTR_SELINUX_SUFFIX "selinux"
#define XATTR_NAME_SELINUX XATTR_SECURITY_PREFIX XATTR_SELINUX_SUFFIX
@@ -2309,6 +2312,7 @@ static int selinux_inode_listsecurity(st
static int selinux_file_permission(struct file *file, int mask)
{
+ int rc;
struct inode *inode = file->f_dentry->d_inode;
if (!mask) {
@@ -2320,8 +2324,12 @@ static int selinux_file_permission(struc
if ((file->f_flags & O_APPEND) && (mask & MAY_WRITE))
mask |= MAY_APPEND;
- return file_has_perm(current, file,
- file_mask_to_av(inode->i_mode, mask));
+ rc = file_has_perm(current, file,
+ file_mask_to_av(inode->i_mode, mask));
+ if (rc)
+ return rc;
+
+ return selinux_netlbl_inode_permission(inode, mask);
}
static int selinux_file_alloc_security(struct file *file)
@@ -2972,8 +2980,8 @@ out:
return err;
}
-static void selinux_socket_post_create(struct socket *sock, int family,
- int type, int protocol, int kern)
+static int selinux_socket_post_create(struct socket *sock, int family,
+ int type, int protocol, int kern)
{
struct inode_security_struct *isec;
struct task_security_struct *tsec;
@@ -2987,7 +2995,7 @@ static void selinux_socket_post_create(s
isec->sid = kern ? SECINITSID_KERNEL : newsid;
isec->initialized = 1;
- return;
+ return selinux_netlbl_socket_create(sock, family, isec->sid);
}
/* Range of port numbers used to automatically bind.
@@ -3162,13 +3170,25 @@ static int selinux_socket_accept(struct
newisec->sid = isec->sid;
newisec->initialized = 1;
- return 0;
+ return selinux_netlbl_socket_accept(sock, newsock);
+}
+
+static void selinux_socket_post_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ selinux_netlbl_socket_post_accept(sock, newsock);
}
static int selinux_socket_sendmsg(struct socket *sock, struct msghdr *msg,
int size)
{
- return socket_has_perm(current, sock, SOCKET__WRITE);
+ int rc;
+
+ rc = socket_has_perm(current, sock, SOCKET__WRITE);
+ if (rc)
+ return rc;
+
+ return selinux_netlbl_inode_permission(SOCK_INODE(sock), MAY_WRITE);
}
static int selinux_socket_recvmsg(struct socket *sock, struct msghdr *msg,
@@ -3376,6 +3396,10 @@ static int selinux_socket_sock_rcv_skb(s
if (err)
goto out;
+ err = selinux_netlbl_sock_rcv_skb(sk, skb, sock_class, sock_sid, &ad);
+ if (err)
+ goto out;
+
err = selinux_xfrm_sock_rcv_skb(sock_sid, skb);
out:
return err;
@@ -3399,8 +3423,10 @@ static int selinux_socket_getpeersec_str
peer_sid = ssec->peer_sid;
}
else if (isec->sclass == SECCLASS_TCP_SOCKET) {
- peer_sid = selinux_socket_getpeer_stream(sock->sk);
-
+ peer_sid = selinux_netlbl_socket_getpeersec_stream(sock);
+ if (peer_sid == SECSID_NULL ||
+ peer_sid == SECINITSID_UNLABELED)
+ peer_sid = selinux_socket_getpeer_stream(sock->sk);
if (peer_sid == SECSID_NULL) {
err = -ENOPROTOOPT;
goto out;
@@ -3441,8 +3467,12 @@ static int selinux_socket_getpeersec_dgr
if (skb->sk->sk_family == PF_UNIX)
selinux_get_inode_sid(SOCK_INODE(skb->sk->sk_socket),
&peer_sid);
- else
- peer_sid = selinux_socket_getpeer_dgram(skb);
+ else {
+ peer_sid = selinux_netlbl_socket_getpeersec_dgram(skb);
+ if (peer_sid == SECSID_NULL ||
+ peer_sid == SECINITSID_UNLABELED)
+ peer_sid = selinux_socket_getpeer_dgram(skb);
+ }
if (peer_sid == SECSID_NULL)
return -EINVAL;
@@ -4505,6 +4535,7 @@ static struct security_operations selinu
.socket_connect = selinux_socket_connect,
.socket_listen = selinux_socket_listen,
.socket_accept = selinux_socket_accept,
+ .socket_post_accept = selinux_socket_post_accept,
.socket_sendmsg = selinux_socket_sendmsg,
.socket_recvmsg = selinux_socket_recvmsg,
.socket_getsockname = selinux_socket_getsockname,
Index: linux-2.6.18-rc1/security/selinux/include/objsec.h
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/include/objsec.h
+++ linux-2.6.18-rc1/security/selinux/include/objsec.h
@@ -37,6 +37,16 @@ struct task_security_struct {
u32 ptrace_sid; /* SID of ptrace parent */
};
+struct netlbl_security_struct {
+ u32 peer_sid; /* SID of the connected peer */
+ enum { /* NetLabel state */
+ NLBL_UNSET = 0,
+ NLBL_ACCEPT,
+ NLBL_REQUIRE,
+ NLBL_LABELED,
+ } state;
+};
+
struct inode_security_struct {
struct inode *inode; /* back pointer to inode object */
struct list_head list; /* list of inode_security_struct */
@@ -46,6 +56,9 @@ struct inode_security_struct {
unsigned char initialized; /* initialization flag */
struct semaphore sem;
unsigned char inherit; /* inherit SID from parent entry */
+#ifdef CONFIG_NETLABEL
+ struct netlbl_security_struct netlbl;
+#endif
};
struct file_security_struct {
Index: linux-2.6.18-rc1/security/selinux/include/selinux_netlabel.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/security/selinux/include/selinux_netlabel.h
@@ -0,0 +1,100 @@
+/*
+ * SELinux interface to the NetLabel subsystem
+ *
+ * Author : Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _SELINUX_NETLABEL_H_
+#define _SELINUX_NETLABEL_H_
+
+#ifdef CONFIG_NETLABEL
+void selinux_netlbl_cache_invalidate(void);
+int selinux_netlbl_inode_permission(struct inode *inode,
+ int mask);
+
+int selinux_netlbl_socket_create(struct socket *sock,
+ const int sock_family,
+ u32 sid);
+int selinux_netlbl_socket_accept(struct socket *sock,
+ struct socket *newsock);
+void selinux_netlbl_socket_post_accept(struct socket *sock,
+ struct socket *newsock);
+int selinux_netlbl_sock_rcv_skb(struct sock *sk,
+ struct sk_buff *skb,
+ const u16 sock_class,
+ const u32 sock_sid,
+ struct avc_audit_data *ad);
+u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock);
+u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb);
+#else
+static inline void selinux_netlbl_cache_invalidate(void)
+{
+ return;
+}
+
+static inline int selinux_netlbl_inode_permission(struct inode *inode,
+ int mask)
+{
+ return 0;
+}
+
+static inline int selinux_netlbl_socket_create(struct socket *sock,
+ const int sock_family,
+ u32 sid)
+{
+ return 0;
+}
+
+static inline int selinux_netlbl_socket_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ return 0;
+}
+
+static inline void selinux_netlbl_socket_post_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ return;
+}
+
+static inline int selinux_netlbl_sock_rcv_skb(struct sock *sk,
+ struct sk_buff *skb,
+ const u16 sock_class,
+ const u32 sock_sid,
+ struct avc_audit_data *ad)
+{
+ return 0;
+}
+
+static inline u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock)
+{
+ return SECSID_NULL;
+}
+
+static inline u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb)
+{
+ return SECSID_NULL;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif
Index: linux-2.6.18-rc1/security/selinux/ss/ebitmap.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/ebitmap.c
+++ linux-2.6.18-rc1/security/selinux/ss/ebitmap.c
@@ -3,6 +3,14 @@
*
* Author : Stephen Smalley, <sds@epoch.ncsc.mil>
*/
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added ebitmap_export() and ebitmap_import()
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
+
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/errno.h>
@@ -59,6 +67,150 @@ int ebitmap_cpy(struct ebitmap *dst, str
return 0;
}
+/**
+ * ebitmap_export - Export an ebitmap to a unsigned char bitmap string
+ * @src: the ebitmap to export
+ * @dst: the resulting bitmap string
+ * @dst_len: length of dst in bytes
+ *
+ * Description:
+ * Allocate a buffer at least src->highbit bits long and export the extensible
+ * bitmap into the buffer. The bitmap string will be in little endian format,
+ * i.e. LSB first. The value returned in dst_len may not the true size of the
+ * buffer as the length of the buffer is rounded up to a multiple of MAPTYPE.
+ * The caller must free the buffer when finished. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int ebitmap_export(const struct ebitmap *src,
+ unsigned char **dst,
+ size_t *dst_len)
+{
+ size_t bitmap_len;
+ unsigned char *bitmap;
+ struct ebitmap_node *iter_node;
+ MAPTYPE node_val;
+ size_t bitmap_byte;
+ unsigned char bitmask;
+
+ if (src == NULL || dst == NULL || dst_len == NULL)
+ return -EINVAL;
+
+ *dst = NULL;
+ *dst_len = 0;
+
+ bitmap_len = src->highbit / 8;
+ if (src->highbit % 8 > 0)
+ bitmap_len += 1;
+ if (bitmap_len == 0)
+ return -EINVAL;
+
+ bitmap = kzalloc(bitmap_len + sizeof(MAPTYPE) -
+ (bitmap_len % sizeof(MAPTYPE)),
+ GFP_ATOMIC);
+ if (bitmap == NULL)
+ return -ENOMEM;
+
+ iter_node = src->node;
+ do {
+ bitmap_byte = iter_node->startbit / 8;
+ bitmask = 0x80;
+ node_val = iter_node->map;
+ do {
+ if (bitmask == 0) {
+ bitmap_byte++;
+ bitmask = 0x80;
+ }
+ if (node_val & (MAPTYPE)0x01)
+ bitmap[bitmap_byte] |= bitmask;
+ node_val >>= 1;
+ bitmask >>= 1;
+ } while (node_val > 0);
+ iter_node = iter_node->next;
+ } while (iter_node);
+
+ *dst = bitmap;
+ *dst_len = bitmap_len;
+ return 0;
+}
+
+/**
+ * ebitmap_import - Import an unsigned char bitmap string into an ebitmap
+ * @src: the bitmap string
+ * @src_len: the bitmap length in bytes
+ * @dst: the empty ebitmap
+ *
+ * Description:
+ * This function takes a little endian bitmap string in src and imports it into
+ * the ebitmap pointed to by dst. Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int ebitmap_import(const unsigned char *src,
+ const size_t src_len,
+ struct ebitmap *dst)
+{
+ size_t src_off = 0;
+ struct ebitmap_node *node_new;
+ struct ebitmap_node *node_last = NULL;
+ size_t src_rem = src_len;
+ MAPTYPE tmp_val;
+ size_t iter;
+ size_t iter_bit;
+ unsigned char src_byte;
+
+ if (src == NULL || dst == NULL || src_len == 0)
+ return -EINVAL;
+
+ do {
+ node_new = kzalloc(sizeof(*node_new), GFP_ATOMIC);
+ if (node_new == NULL) {
+ ebitmap_destroy(dst);
+ return -ENOMEM;
+ }
+
+ if (src_rem >= sizeof(MAPTYPE))
+ iter = sizeof(MAPTYPE);
+ else
+ iter = src_rem;
+ tmp_val = 0;
+ while (iter > 0) {
+ src_byte = src[src_off + --iter];
+ if (src_byte > 0)
+ for (iter_bit = 0; iter_bit < 8; iter_bit++) {
+ tmp_val <<= 1;
+ tmp_val |= src_byte & 0x01;
+ src_byte >>= 1;
+ }
+ else
+ tmp_val <<= 8;
+ }
+ node_new->map = tmp_val;
+ node_new->startbit = src_off * 8;
+
+ if (node_last != NULL)
+ node_last->next = node_new;
+ else
+ dst->node = node_new;
+ node_last = node_new;
+
+ if (src_rem >= sizeof(MAPTYPE)) {
+ src_off += sizeof(MAPTYPE);
+ src_rem -= sizeof(MAPTYPE);
+ } else
+ src_off += src_rem;
+ } while (src_off < src_len);
+
+ tmp_val = node_last->map;
+ dst->highbit = node_last->startbit;
+ while (tmp_val >= 1) {
+ dst->highbit += 1;
+ tmp_val >>= 1;
+ }
+
+ return 0;
+}
+
int ebitmap_contains(struct ebitmap *e1, struct ebitmap *e2)
{
struct ebitmap_node *n1, *n2;
Index: linux-2.6.18-rc1/security/selinux/ss/ebitmap.h
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/ebitmap.h
+++ linux-2.6.18-rc1/security/selinux/ss/ebitmap.h
@@ -69,6 +69,12 @@ static inline int ebitmap_node_get_bit(s
int ebitmap_cmp(struct ebitmap *e1, struct ebitmap *e2);
int ebitmap_cpy(struct ebitmap *dst, struct ebitmap *src);
+int ebitmap_export(const struct ebitmap *src,
+ unsigned char **dst,
+ size_t *dst_len);
+int ebitmap_import(const unsigned char *src,
+ const size_t src_len,
+ struct ebitmap *dst);
int ebitmap_contains(struct ebitmap *e1, struct ebitmap *e2);
int ebitmap_get_bit(struct ebitmap *e, unsigned long bit);
int ebitmap_set_bit(struct ebitmap *e, unsigned long bit, int value);
Index: linux-2.6.18-rc1/security/selinux/ss/mls.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/mls.c
+++ linux-2.6.18-rc1/security/selinux/ss/mls.c
@@ -10,6 +10,13 @@
*
* Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
*/
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added support to import/export the MLS label
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
#include <linux/kernel.h>
#include <linux/slab.h>
@@ -585,3 +592,148 @@ int mls_compute_sid(struct context *scon
return -EINVAL;
}
+/**
+ * mls_export_lvl - Export the MLS sensitivity levels
+ * @context: the security context
+ * @low: the low sensitivity level
+ * @high: the high sensitivity level
+ *
+ * Description:
+ * Given the security context copy the low MLS sensitivity level into lvl_low
+ * and the high sensitivity level in lvl_high. The MLS levels are only
+ * exported if the pointers are not NULL, if they are NULL then that level is
+ * not exported.
+ *
+ */
+void mls_export_lvl(const struct context *context, u32 *low, u32 *high)
+{
+ if (!selinux_mls_enabled)
+ return;
+
+ if (low != NULL)
+ *low = context->range.level[0].sens - 1;
+ if (high != NULL)
+ *high = context->range.level[1].sens - 1;
+}
+
+/**
+ * mls_import_lvl - Import the MLS sensitivity levels
+ * @context: the security context
+ * @low: the low sensitivity level
+ * @high: the high sensitivity level
+ *
+ * Description:
+ * Given the security context and the two sensitivty levels, set the MLS levels
+ * in the context according the two given as parameters. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+void mls_import_lvl(struct context *context, const u32 low, const u32 high)
+{
+ if (!selinux_mls_enabled)
+ return;
+
+ context->range.level[0].sens = low + 1;
+ context->range.level[1].sens = high + 1;
+}
+
+/**
+ * mls_export_cat - Export the MLS categories
+ * @context: the security context
+ * @low: the low category
+ * @low_len: length of the cat_low bitmap in bytes
+ * @high: the high category
+ * @high_len: length of the cat_high bitmap in bytes
+ *
+ * Description:
+ * Given the security context export the low MLS category bitmap into cat_low
+ * and the high category bitmap into cat_high. The MLS categories are only
+ * exported if the pointers are not NULL, if they are NULL then that level is
+ * not exported. The caller is responsibile for freeing the memory when
+ * finished. Returns zero on success, negative values on failure.
+ *
+ */
+int mls_export_cat(const struct context *context,
+ unsigned char **low,
+ size_t *low_len,
+ unsigned char **high,
+ size_t *high_len)
+{
+ int rc = -EPERM;
+
+ if (!selinux_mls_enabled)
+ return 0;
+
+ if (low != NULL) {
+ rc = ebitmap_export(&context->range.level[0].cat,
+ low,
+ low_len);
+ if (rc != 0)
+ goto export_cat_failure;
+ }
+ if (high != NULL) {
+ rc = ebitmap_export(&context->range.level[1].cat,
+ high,
+ high_len);
+ if (rc != 0)
+ goto export_cat_failure;
+ }
+
+ return 0;
+
+export_cat_failure:
+ if (low != NULL)
+ kfree(*low);
+ if (high != NULL)
+ kfree(*high);
+ return rc;
+}
+
+/**
+ * mls_import_cat - Import the MLS categories
+ * @context: the security context
+ * @low: the low category
+ * @low_len: length of the cat_low bitmap in bytes
+ * @high: the high category
+ * @high_len: length of the cat_high bitmap in bytes
+ *
+ * Description:
+ * Given the security context and the two category bitmap strings import the
+ * categories into the security context. The MLS categories are only imported
+ * if the pointers are not NULL, if they are NULL they are skipped. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+int mls_import_cat(struct context *context,
+ const unsigned char *low,
+ const size_t low_len,
+ const unsigned char *high,
+ const size_t high_len)
+{
+ int rc = -EPERM;
+
+ if (!selinux_mls_enabled)
+ return 0;
+
+ if (low != NULL) {
+ rc = ebitmap_import(low,
+ low_len,
+ &context->range.level[0].cat);
+ if (rc != 0)
+ goto import_cat_failure;
+ }
+ if (high != NULL) {
+ rc = ebitmap_import(high,
+ high_len,
+ &context->range.level[1].cat);
+ if (rc != 0)
+ goto import_cat_failure;
+ }
+
+ return 0;
+
+import_cat_failure:
+ ebitmap_destroy(&context->range.level[0].cat);
+ ebitmap_destroy(&context->range.level[1].cat);
+ return rc;
+}
Index: linux-2.6.18-rc1/security/selinux/ss/mls.h
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/mls.h
+++ linux-2.6.18-rc1/security/selinux/ss/mls.h
@@ -10,6 +10,13 @@
*
* Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
*/
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added support to import/export the MLS label
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
#ifndef _SS_MLS_H_
#define _SS_MLS_H_
@@ -42,5 +49,19 @@ int mls_compute_sid(struct context *scon
int mls_setup_user_range(struct context *fromcon, struct user_datum *user,
struct context *usercon);
+void mls_export_lvl(const struct context *context, u32 *low, u32 *high);
+void mls_import_lvl(struct context *context, const u32 low, const u32 high);
+
+int mls_export_cat(const struct context *context,
+ unsigned char **low,
+ size_t *low_len,
+ unsigned char **high,
+ size_t *high_len);
+int mls_import_cat(struct context *context,
+ const unsigned char *low,
+ const size_t low_len,
+ const unsigned char *high,
+ const size_t high_len);
+
#endif /* _SS_MLS_H */
Index: linux-2.6.18-rc1/security/selinux/ss/services.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/services.c
+++ linux-2.6.18-rc1/security/selinux/ss/services.c
@@ -13,6 +13,11 @@
*
* Added conditional policy language extensions
*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added support for NetLabel
+ *
+ * Copyright (C) 2006 Hewlett-Packard Development Company, L.P.
* Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
* Copyright (C) 2003 - 2004 Tresys Technology, LLC
* Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
@@ -29,6 +34,8 @@
#include <linux/sched.h>
#include <linux/audit.h>
#include <linux/mutex.h>
+#include <net/sock.h>
+#include <net/netlabel.h>
#include "flask.h"
#include "avc.h"
@@ -40,6 +47,8 @@
#include "services.h"
#include "conditional.h"
#include "mls.h"
+#include "objsec.h"
+#include "selinux_netlabel.h"
extern void selnl_notify_policyload(u32 seqno);
unsigned int policydb_loaded_version;
@@ -1241,6 +1250,7 @@ int security_load_policy(void *data, siz
selinux_complete_init();
avc_ss_reset(seqno);
selnl_notify_policyload(seqno);
+ selinux_netlbl_cache_invalidate();
return 0;
}
@@ -1295,6 +1305,7 @@ int security_load_policy(void *data, siz
avc_ss_reset(seqno);
selnl_notify_policyload(seqno);
+ selinux_netlbl_cache_invalidate();
return 0;
@@ -2064,3 +2075,547 @@ void selinux_audit_set_callback(int (*ca
{
aurule_callback = callback;
}
+
+#ifdef CONFIG_NETLABEL
+/*
+ * This is the structure we store inside the NetLabel cache block.
+ */
+#define NETLBL_CACHE(x) ((struct netlbl_cache *)(x))
+#define NETLBL_CACHE_T_NONE 0
+#define NETLBL_CACHE_T_SID 1
+#define NETLBL_CACHE_T_MLSLBL 2
+struct netlbl_cache {
+ u32 type;
+ union {
+ u32 sid;
+ struct mls_level mls_label;
+ } data;
+};
+
+/**
+ * selinux_netlbl_cache_free - Free the NetLabel cached data
+ * @data: the data to free
+ *
+ * Description:
+ * This function is intended to be used as the free() callback inside the
+ * netlbl_lsm_cache structure.
+ *
+ */
+static void selinux_netlbl_cache_free(const void *data)
+{
+ struct netlbl_cache *cache = NETLBL_CACHE(data);
+ switch (cache->type) {
+ case NETLBL_CACHE_T_MLSLBL:
+ ebitmap_destroy(&cache->data.mls_label.cat);
+ break;
+ }
+ kfree(data);
+}
+
+/**
+ * selinux_netlbl_cache_add - Add an entry to the NetLabel cache
+ * @skb: the packet
+ * @ctx: the SELinux context
+ *
+ * Description:
+ * Attempt to cache the context in @ctx, which was derived from the packet in
+ * @skb, in the NetLabel subsystem cache.
+ *
+ */
+static void selinux_netlbl_cache_add(struct sk_buff *skb, struct context *ctx)
+{
+ struct netlbl_cache *cache = NULL;
+ struct netlbl_lsm_secattr secattr;
+ u32 cache_clear = 1;
+
+ netlbl_secattr_init(&secattr);
+
+ cache = kzalloc(sizeof(*cache), GFP_ATOMIC);
+ if (cache == NULL)
+ goto netlbl_cache_add_return;
+ secattr.cache.free = selinux_netlbl_cache_free;
+ secattr.cache.data = (void *)cache;
+
+ cache->type = NETLBL_CACHE_T_MLSLBL;
+ if (ebitmap_cpy(&cache->data.mls_label.cat,
+ &ctx->range.level[0].cat) != 0)
+ goto netlbl_cache_add_return;
+ cache->data.mls_label.sens = ctx->range.level[0].sens;
+
+ if (netlbl_cache_add(skb, &secattr) != 0)
+ goto netlbl_cache_add_return;
+ cache_clear = 0;
+
+netlbl_cache_add_return:
+ netlbl_secattr_destroy(&secattr, cache_clear);
+}
+
+/**
+ * selinux_netlbl_cache_invalidate - Invalidate the NetLabel cache
+ *
+ * Description:
+ * Invalidate the NetLabel security attribute mapping cache.
+ *
+ */
+void selinux_netlbl_cache_invalidate(void)
+{
+ netlbl_cache_invalidate();
+}
+
+/**
+ * selinux_netlbl_secattr_to_sid - Convert a NetLabel secattr to a SELinux SID
+ * @skb: the network packet
+ * @secattr: the NetLabel packet security attributes
+ * @base_sid: the SELinux SID to use as a context for MLS only attributes
+ * @sid: the SELinux SID
+ *
+ * Description:
+ * Convert the given NetLabel packet security attributes in @secattr into a
+ * SELinux SID. If the @secattr field does not contain a full SELinux
+ * SID/context then use the context in @base_sid as the foundation. If @skb
+ * is not NULL attempt to cache as much data as possibile. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_secattr_to_sid(struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr,
+ const u32 base_sid,
+ u32 *sid)
+{
+ int rc = -EIDRM;
+ struct context *ctx;
+ struct context ctx_new;
+ struct netlbl_cache *cache;
+ u32 ctx_new_destroy = 0;
+
+ POLICY_RDLOCK;
+
+ if (secattr->cache.data) {
+ cache = NETLBL_CACHE(secattr->cache.data);
+ switch (cache->type) {
+ case NETLBL_CACHE_T_SID:
+ *sid = cache->data.sid;
+ break;
+ case NETLBL_CACHE_T_MLSLBL:
+ ctx = sidtab_search(&sidtab, base_sid);
+ if (ctx == NULL)
+ goto netlbl_secattr_to_sid_return;
+ rc = context_cpy(&ctx_new, ctx);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+ ctx_new_destroy = 1;
+ mls_context_destroy(&ctx_new);
+
+ ctx_new.range.level[0].sens =
+ cache->data.mls_label.sens;
+ rc = ebitmap_cpy(&ctx_new.range.level[0].cat,
+ &cache->data.mls_label.cat);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+ ctx_new.range.level[1].sens =
+ cache->data.mls_label.sens;
+ rc = ebitmap_cpy(&ctx_new.range.level[1].cat,
+ &cache->data.mls_label.cat);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+
+ rc = sidtab_context_to_sid(&sidtab, &ctx_new, sid);
+ break;
+ default:
+ goto netlbl_secattr_to_sid_return;
+ }
+ } else if (secattr->mls_lvl_vld) {
+ ctx = sidtab_search(&sidtab, base_sid);
+ if (ctx == NULL)
+ goto netlbl_secattr_to_sid_return;
+ rc = context_cpy(&ctx_new, ctx);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+ ctx_new_destroy = 1;
+ mls_context_destroy(&ctx_new);
+
+ mls_import_lvl(&ctx_new, secattr->mls_lvl, secattr->mls_lvl);
+ if (secattr->mls_cat) {
+ if (mls_import_cat(&ctx_new,
+ secattr->mls_cat,
+ secattr->mls_cat_len,
+ secattr->mls_cat,
+ secattr->mls_cat_len) != 0)
+ goto netlbl_secattr_to_sid_return;
+ }
+
+ rc = mls_context_isvalid(&policydb, &ctx_new);
+ if (rc != 1)
+ goto netlbl_secattr_to_sid_return;
+
+ if (skb != NULL)
+ selinux_netlbl_cache_add(skb, &ctx_new);
+
+ rc = sidtab_context_to_sid(&sidtab, &ctx_new, sid);
+ } else
+ *sid = SECINITSID_UNLABELED;
+
+ rc = 0;
+
+netlbl_secattr_to_sid_return:
+ POLICY_RDUNLOCK;
+ if (ctx_new_destroy)
+ context_destroy(&ctx_new);
+ return rc;
+}
+
+/**
+ * selinux_netlbl_skbuff_getsid - Get the sid of a packet using NetLabel
+ * @skb: the packet
+ * @base_sid: the SELinux SID to use as a context for MLS only attributes
+ * @sid: the SID
+ *
+ * Description:
+ * Call the NetLabel mechanism to get the security attributes of the given
+ * packet and use those attributes to determine the correct context/SID to
+ * assign to the packet. Returns zero on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_skbuff_getsid(struct sk_buff *skb,
+ const u32 base_sid,
+ u32 *sid)
+{
+ int rc;
+ struct netlbl_lsm_secattr secattr;
+
+ netlbl_secattr_init(&secattr);
+ rc = netlbl_skbuff_getattr(skb, &secattr);
+ if (rc == 0)
+ rc = selinux_netlbl_secattr_to_sid(skb,
+ &secattr,
+ base_sid,
+ sid);
+ netlbl_secattr_destroy(&secattr, 0);
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_setsid - Label a socket using the NetLabel mechanism
+ * @sock: the socket to label
+ * @sock_family: the socket family
+ * @sid: the SID to use
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism using the given
+ * SID. Returns zero values on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_socket_setsid(struct socket *sock,
+ const int sock_family,
+ u32 sid)
+{
+ int rc;
+ struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+ struct netlbl_lsm_secattr secattr;
+ struct context *ctx;
+
+ if (!ss_initialized)
+ return 0;
+
+ POLICY_RDLOCK;
+
+ ctx = sidtab_search(&sidtab, sid);
+ if (ctx != NULL) {
+ netlbl_secattr_init(&secattr);
+ secattr.domain = kstrdup(
+ policydb.p_type_val_to_name[ctx->type - 1],
+ GFP_ATOMIC);
+ mls_export_lvl(ctx, &secattr.mls_lvl, NULL);
+ secattr.mls_lvl_vld = 1;
+ mls_export_cat(ctx,
+ &secattr.mls_cat,
+ &secattr.mls_cat_len,
+ NULL,
+ NULL);
+
+ rc = netlbl_socket_setattr(sock, &secattr);
+ if (rc == 0)
+ isec->netlbl.state = NLBL_LABELED;
+
+ netlbl_secattr_destroy(&secattr, 0);
+ } else
+ rc = -ENOENT;
+
+ POLICY_RDUNLOCK;
+
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_peeksid - Get the SID of the first queued packet
+ * @sock: the socket to query
+ * @sid: the packet's SID
+ *
+ * Description:
+ * Examine the first incoming packet in the socket's queue and determine the
+ * packet's SELinux SID. Return zero on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_socket_peeksid(struct socket *sock,
+ u32 *sid)
+{
+ int rc;
+ struct netlbl_lsm_secattr secattr;
+ struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+
+ netlbl_secattr_init(&secattr);
+ rc = netlbl_socket_peekattr(sock, &secattr);
+ if (rc == 0)
+ rc = selinux_netlbl_secattr_to_sid(NULL,
+ &secattr,
+ isec->sid,
+ sid);
+ netlbl_secattr_destroy(&secattr, 0);
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_create - Label a socket using the NetLabel mechanism
+ * @sock: the socket to label
+ * @sock_family: the socket family
+ * @sid: the SID to use
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism using the given
+ * SID. Returns zero values on success, negative values on failure.
+ *
+ */
+int selinux_netlbl_socket_create(struct socket *sock,
+ const int sock_family,
+ u32 sid)
+{
+ struct inode_security_struct *isec;
+
+ if (sock_family != PF_INET)
+ return 0;
+
+ isec = SOCK_INODE(sock)->i_security;
+ isec->netlbl.state = NLBL_REQUIRE;
+ return selinux_netlbl_socket_setsid(sock, sock_family, sid);
+}
+
+/**
+ * selinux_netlbl_socket_accept - NetLabel accept hook
+ * @sock: the original socket
+ * @newsock: the new accept()ed socket
+ *
+ * Description:
+ * Mark @newsock as requiring a NetLabel label and do setup required for
+ * further processing by the post_accept and sock_rcv_skb hooks. Returns zero
+ * on success, negative values on failure.
+ *
+ */
+int selinux_netlbl_socket_accept(struct socket *sock, struct socket *newsock)
+{
+ struct inode_security_struct *isec;
+ struct inode_security_struct *newisec;
+
+ isec = SOCK_INODE(sock)->i_security;
+ if (isec->netlbl.state == NLBL_UNSET)
+ return 0;
+
+ newisec = SOCK_INODE(newsock)->i_security;
+ newisec->netlbl.state = NLBL_ACCEPT;
+
+ return 0;
+}
+
+/**
+ * selinux_netlbl_socket_post_accept - NetLabel post_accept hook
+ * @sock: the original socket
+ * @newsock: the new accept()ed socket
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism based on the
+ * connection and the original/parent socket's SID.
+ *
+ */
+void selinux_netlbl_socket_post_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ struct inode_security_struct *newisec;
+ struct netlbl_lsm_secattr secattr;
+
+ newisec = SOCK_INODE(newsock)->i_security;
+ if (newisec->netlbl.state != NLBL_ACCEPT)
+ return;
+ netlbl_secattr_init(&secattr);
+ if (netlbl_socket_getattr(newsock, &secattr) != 0)
+ goto netlbl_socket_post_accept_return;
+ newisec = SOCK_INODE(newsock)->i_security;
+ if (selinux_netlbl_secattr_to_sid(NULL,
+ &secattr,
+ newisec->sid,
+ &newisec->netlbl.peer_sid) != 0)
+ goto netlbl_socket_post_accept_return;
+ newisec->sid = newisec->netlbl.peer_sid;
+
+ /* XXX - try to set the NetLabel here to try and save some time in
+ selinux_netlbl_inode_permission() later */
+ selinux_netlbl_socket_setsid(newsock,
+ newsock->sk->sk_family,
+ newisec->sid);
+
+netlbl_socket_post_accept_return:
+ if (newisec->netlbl.state != NLBL_LABELED)
+ newisec->netlbl.state = NLBL_REQUIRE;
+ netlbl_secattr_destroy(&secattr, 0);
+}
+
+/**
+ * selinux_netlbl_inode_permission - Verify the file/socket is NetLabel labeled
+ * @inode: the file descriptor's inode
+ * @mask: the permission mask
+ *
+ * Description:
+ * Looks at a file's inode and if it is marked as a socket protected by
+ * NetLabel then verify that the socket has been labeled, if not try to label
+ * the socket now with the task's SID. Returns zero on success, negative
+ * values on failure.
+ *
+ */
+int selinux_netlbl_inode_permission(struct inode *inode, int mask)
+{
+ int rc = 0;
+ struct inode_security_struct *isec = inode->i_security;
+ struct netlbl_security_struct *nlblsec = &isec->netlbl;
+ struct socket *sock;
+
+ down(&isec->sem);
+ if (nlblsec->state == NLBL_REQUIRE &&
+ (mask & (MAY_WRITE | MAY_APPEND))) {
+ sock = SOCKET_I(inode);
+ rc = selinux_netlbl_socket_setsid(sock,
+ sock->sk->sk_family,
+ isec->sid);
+ }
+ up(&isec->sem);
+
+ return rc;
+}
+
+/**
+ * selinux_netlbl_sock_rcv_skb - Do an inbound access check using NetLabel
+ * @sk: the packet's sock
+ * @skb: the packet
+ * @sock_class: the socket class
+ * @sock_sid: the socket's SID
+ * @ad: the audit data
+ *
+ * Description:
+ * Fetch the NetLabel security attributes from @skb and perform an access check
+ * against the receiving socket. Returns zero on success, negative values on
+ * error.
+ *
+ */
+int selinux_netlbl_sock_rcv_skb(struct sock *sk,
+ struct sk_buff *skb,
+ const u16 sock_class,
+ const u32 sock_sid,
+ struct avc_audit_data *ad)
+{
+ int rc;
+ u32 netlbl_sid;
+ u32 recv_perm;
+
+ rc = selinux_netlbl_skbuff_getsid(skb, sock_sid, &netlbl_sid);
+ if (rc != 0)
+ return rc;
+
+ if (netlbl_sid != SECINITSID_UNLABELED) {
+ switch (sock_class) {
+ case SECCLASS_UDP_SOCKET:
+ recv_perm = UDP_SOCKET__RECV_MSG;
+ break;
+ case SECCLASS_TCP_SOCKET:
+ recv_perm = TCP_SOCKET__RECV_MSG;
+ break;
+ default:
+ recv_perm = RAWIP_SOCKET__RECV_MSG;
+ break;
+ }
+
+ rc = avc_has_perm(sock_sid,
+ netlbl_sid,
+ sock_class,
+ recv_perm,
+ ad);
+ if (rc != 0) {
+ struct inode_security_struct *isec;
+ u32 netlbl_send_err = 1;
+
+ read_lock_bh(&sk->sk_callback_lock);
+ isec = SOCK_INODE(sk->sk_socket)->i_security;
+ if (isec->netlbl.state == NLBL_ACCEPT)
+ netlbl_send_err = 0;
+ read_unlock_bh(&sk->sk_callback_lock);
+
+ if (netlbl_send_err)
+ netlbl_skbuff_err(skb, rc);
+ }
+ }
+
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_peersid - Return the peer SID of a connected socket
+ * @sock: the socket
+ *
+ * Description:
+ * Examine @sock to find the connected peer's SID, if that is not possibile
+ * check the socket's receive queue and take the SID from the incoming
+ * packets. Returns the SID on success or SECSID_NULL on error.
+ *
+ */
+u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock)
+{
+ u32 peer_sid;
+ struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+
+ if (isec->netlbl.peer_sid != 0)
+ peer_sid = isec->netlbl.peer_sid;
+ else {
+ if (selinux_netlbl_socket_peeksid(sock, &peer_sid) != 0)
+ peer_sid = SECSID_NULL;
+ }
+
+ return peer_sid;
+}
+
+/**
+ * selinux_netlbl_socket_getpeersec_dgram - Return the SID of a NetLabel packet
+ * @skb: the packet
+ *
+ * Description:
+ * Examine @skb to find the SID assigned to it by NetLabel. Returns the SID on
+ * success, SECSID_NULL on error.
+ *
+ */
+u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb)
+{
+ int peer_sid;
+ struct sock *sk = skb->sk;
+ struct inode_security_struct *isec;
+
+ if (sk == NULL)
+ return SECSID_NULL;
+
+ read_lock_bh(&skb->sk->sk_callback_lock);
+ if (skb->sk->sk_socket == NULL) {
+ read_unlock_bh(&skb->sk->sk_callback_lock);
+ return SECSID_NULL;
+ }
+ isec = SOCK_INODE(skb->sk->sk_socket)->i_security;
+ if (selinux_netlbl_skbuff_getsid(skb, isec->sid, &peer_sid) != 0)
+ peer_sid = SECSID_NULL;
+ read_unlock_bh(&sk->sk_callback_lock);
+
+ return peer_sid;
+}
+#endif /* CONFIG_NETLABEL */
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 6/7] NetLabel: SELinux support
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Add NetLabel support to the SELinux LSM and modify the socket_post_create() LSM
hook to return an error code. The most significant part of this patch is the
addition of NetLabel hooks into the following SELinux LSM hooks:
* selinux_file_permission()
* selinux_socket_sendmsg()
* selinux_socket_post_create()
* selinux_socket_accept()
* selinux_socket_post_accept() [NEW]
* selinux_socket_sock_rcv_skb()
* selinux_socket_getpeersec_stream()
* selinux_socket_getpeersec_dgram()
The basic reasoning behind this patch is that outgoing packets are "NetLabel'd"
by labeling their socket and the NetLabel security attributes are checked via
the additional hook in selinux_socket_sock_rcv_skb(). NetLabel itself is only
a labeling mechanism, similar to filesystem extended attributes, it is up to
the SELinux enforcement mechanism to perform the actual access checks.
In addition to the changes outlined above this patch also includes some changes
to the extended bitmap (ebitmap) and multi-level security (mls) code to import
and export SELinux TE/MLS attributes into and out of NetLabel.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
include/linux/security.h | 25 -
net/socket.c | 17
security/dummy.c | 6
security/selinux/hooks.c | 53 ++
security/selinux/include/objsec.h | 13
security/selinux/include/selinux_netlabel.h | 100 +++++
security/selinux/ss/ebitmap.c | 152 +++++++
security/selinux/ss/ebitmap.h | 6
security/selinux/ss/mls.c | 152 +++++++
security/selinux/ss/mls.h | 21 +
security/selinux/ss/services.c | 555 ++++++++++++++++++++++++++++
11 files changed, 1070 insertions(+), 30 deletions(-)
Index: linux-2.6.18-rc1/include/linux/security.h
===================================================================
--- linux-2.6.18-rc1.orig/include/linux/security.h
+++ linux-2.6.18-rc1/include/linux/security.h
@@ -1296,8 +1296,8 @@ struct security_operations {
int (*unix_may_send) (struct socket * sock, struct socket * other);
int (*socket_create) (int family, int type, int protocol, int kern);
- void (*socket_post_create) (struct socket * sock, int family,
- int type, int protocol, int kern);
+ int (*socket_post_create) (struct socket * sock, int family,
+ int type, int protocol, int kern);
int (*socket_bind) (struct socket * sock,
struct sockaddr * address, int addrlen);
int (*socket_connect) (struct socket * sock,
@@ -2748,13 +2748,13 @@ static inline int security_socket_create
return security_ops->socket_create(family, type, protocol, kern);
}
-static inline void security_socket_post_create(struct socket * sock,
- int family,
- int type,
- int protocol, int kern)
+static inline int security_socket_post_create(struct socket * sock,
+ int family,
+ int type,
+ int protocol, int kern)
{
- security_ops->socket_post_create(sock, family, type,
- protocol, kern);
+ return security_ops->socket_post_create(sock, family, type,
+ protocol, kern);
}
static inline int security_socket_bind(struct socket * sock,
@@ -2880,11 +2880,12 @@ static inline int security_socket_create
return 0;
}
-static inline void security_socket_post_create(struct socket * sock,
- int family,
- int type,
- int protocol, int kern)
+static inline int security_socket_post_create(struct socket * sock,
+ int family,
+ int type,
+ int protocol, int kern)
{
+ return 0;
}
static inline int security_socket_bind(struct socket * sock,
Index: linux-2.6.18-rc1/net/socket.c
===================================================================
--- linux-2.6.18-rc1.orig/net/socket.c
+++ linux-2.6.18-rc1/net/socket.c
@@ -973,11 +973,18 @@ int sock_create_lite(int family, int typ
goto out;
}
- security_socket_post_create(sock, family, type, protocol, 1);
sock->type = type;
+ err = security_socket_post_create(sock, family, type, protocol, 1);
+ if (err)
+ goto out_release;
+
out:
*res = sock;
return err;
+out_release:
+ sock_release(sock);
+ sock = NULL;
+ goto out;
}
/* No kernel lock held - perfect */
@@ -1213,7 +1220,9 @@ static int __sock_create(int family, int
*/
module_put(net_families[family]->owner);
*res = sock;
- security_socket_post_create(sock, family, type, protocol, kern);
+ err = security_socket_post_create(sock, family, type, protocol, kern);
+ if (err)
+ goto out_release;
out:
net_family_read_unlock();
@@ -1443,13 +1452,13 @@ asmlinkage long sys_accept(int fd, struc
goto out_fd;
}
+ security_socket_post_accept(sock, newsock);
+
/* File flags are not inherited via accept() unlike another OSes. */
fd_install(newfd, newfile);
err = newfd;
- security_socket_post_accept(sock, newsock);
-
out_put:
fput_light(sock->file, fput_needed);
out:
Index: linux-2.6.18-rc1/security/dummy.c
===================================================================
--- linux-2.6.18-rc1.orig/security/dummy.c
+++ linux-2.6.18-rc1/security/dummy.c
@@ -709,10 +709,10 @@ static int dummy_socket_create (int fami
return 0;
}
-static void dummy_socket_post_create (struct socket *sock, int family, int type,
- int protocol, int kern)
+static int dummy_socket_post_create (struct socket *sock, int family, int type,
+ int protocol, int kern)
{
- return;
+ return 0;
}
static int dummy_socket_bind (struct socket *sock, struct sockaddr *address,
Index: linux-2.6.18-rc1/security/selinux/hooks.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/hooks.c
+++ linux-2.6.18-rc1/security/selinux/hooks.c
@@ -12,6 +12,8 @@
* Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
* Copyright (C) 2004-2005 Trusted Computer Solutions, Inc.
* <dgoeddel@trustedcs.com>
+ * Copyright (C) 2006 Hewlett-Packard Development Company, L.P.
+ * Paul Moore, <paul.moore@hp.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2,
@@ -74,6 +76,7 @@
#include "objsec.h"
#include "netif.h"
#include "xfrm.h"
+#include "selinux_netlabel.h"
#define XATTR_SELINUX_SUFFIX "selinux"
#define XATTR_NAME_SELINUX XATTR_SECURITY_PREFIX XATTR_SELINUX_SUFFIX
@@ -2309,6 +2312,7 @@ static int selinux_inode_listsecurity(st
static int selinux_file_permission(struct file *file, int mask)
{
+ int rc;
struct inode *inode = file->f_dentry->d_inode;
if (!mask) {
@@ -2320,8 +2324,12 @@ static int selinux_file_permission(struc
if ((file->f_flags & O_APPEND) && (mask & MAY_WRITE))
mask |= MAY_APPEND;
- return file_has_perm(current, file,
- file_mask_to_av(inode->i_mode, mask));
+ rc = file_has_perm(current, file,
+ file_mask_to_av(inode->i_mode, mask));
+ if (rc)
+ return rc;
+
+ return selinux_netlbl_inode_permission(inode, mask);
}
static int selinux_file_alloc_security(struct file *file)
@@ -2972,8 +2980,8 @@ out:
return err;
}
-static void selinux_socket_post_create(struct socket *sock, int family,
- int type, int protocol, int kern)
+static int selinux_socket_post_create(struct socket *sock, int family,
+ int type, int protocol, int kern)
{
struct inode_security_struct *isec;
struct task_security_struct *tsec;
@@ -2987,7 +2995,7 @@ static void selinux_socket_post_create(s
isec->sid = kern ? SECINITSID_KERNEL : newsid;
isec->initialized = 1;
- return;
+ return selinux_netlbl_socket_create(sock, family, isec->sid);
}
/* Range of port numbers used to automatically bind.
@@ -3162,13 +3170,25 @@ static int selinux_socket_accept(struct
newisec->sid = isec->sid;
newisec->initialized = 1;
- return 0;
+ return selinux_netlbl_socket_accept(sock, newsock);
+}
+
+static void selinux_socket_post_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ selinux_netlbl_socket_post_accept(sock, newsock);
}
static int selinux_socket_sendmsg(struct socket *sock, struct msghdr *msg,
int size)
{
- return socket_has_perm(current, sock, SOCKET__WRITE);
+ int rc;
+
+ rc = socket_has_perm(current, sock, SOCKET__WRITE);
+ if (rc)
+ return rc;
+
+ return selinux_netlbl_inode_permission(SOCK_INODE(sock), MAY_WRITE);
}
static int selinux_socket_recvmsg(struct socket *sock, struct msghdr *msg,
@@ -3376,6 +3396,10 @@ static int selinux_socket_sock_rcv_skb(s
if (err)
goto out;
+ err = selinux_netlbl_sock_rcv_skb(sk, skb, sock_class, sock_sid, &ad);
+ if (err)
+ goto out;
+
err = selinux_xfrm_sock_rcv_skb(sock_sid, skb);
out:
return err;
@@ -3399,8 +3423,10 @@ static int selinux_socket_getpeersec_str
peer_sid = ssec->peer_sid;
}
else if (isec->sclass == SECCLASS_TCP_SOCKET) {
- peer_sid = selinux_socket_getpeer_stream(sock->sk);
-
+ peer_sid = selinux_netlbl_socket_getpeersec_stream(sock);
+ if (peer_sid == SECSID_NULL ||
+ peer_sid == SECINITSID_UNLABELED)
+ peer_sid = selinux_socket_getpeer_stream(sock->sk);
if (peer_sid == SECSID_NULL) {
err = -ENOPROTOOPT;
goto out;
@@ -3441,8 +3467,12 @@ static int selinux_socket_getpeersec_dgr
if (skb->sk->sk_family == PF_UNIX)
selinux_get_inode_sid(SOCK_INODE(skb->sk->sk_socket),
&peer_sid);
- else
- peer_sid = selinux_socket_getpeer_dgram(skb);
+ else {
+ peer_sid = selinux_netlbl_socket_getpeersec_dgram(skb);
+ if (peer_sid == SECSID_NULL ||
+ peer_sid == SECINITSID_UNLABELED)
+ peer_sid = selinux_socket_getpeer_dgram(skb);
+ }
if (peer_sid == SECSID_NULL)
return -EINVAL;
@@ -4505,6 +4535,7 @@ static struct security_operations selinu
.socket_connect = selinux_socket_connect,
.socket_listen = selinux_socket_listen,
.socket_accept = selinux_socket_accept,
+ .socket_post_accept = selinux_socket_post_accept,
.socket_sendmsg = selinux_socket_sendmsg,
.socket_recvmsg = selinux_socket_recvmsg,
.socket_getsockname = selinux_socket_getsockname,
Index: linux-2.6.18-rc1/security/selinux/include/objsec.h
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/include/objsec.h
+++ linux-2.6.18-rc1/security/selinux/include/objsec.h
@@ -37,6 +37,16 @@ struct task_security_struct {
u32 ptrace_sid; /* SID of ptrace parent */
};
+struct netlbl_security_struct {
+ u32 peer_sid; /* SID of the connected peer */
+ enum { /* NetLabel state */
+ NLBL_UNSET = 0,
+ NLBL_ACCEPT,
+ NLBL_REQUIRE,
+ NLBL_LABELED,
+ } state;
+};
+
struct inode_security_struct {
struct inode *inode; /* back pointer to inode object */
struct list_head list; /* list of inode_security_struct */
@@ -46,6 +56,9 @@ struct inode_security_struct {
unsigned char initialized; /* initialization flag */
struct semaphore sem;
unsigned char inherit; /* inherit SID from parent entry */
+#ifdef CONFIG_NETLABEL
+ struct netlbl_security_struct netlbl;
+#endif
};
struct file_security_struct {
Index: linux-2.6.18-rc1/security/selinux/include/selinux_netlabel.h
===================================================================
--- /dev/null
+++ linux-2.6.18-rc1/security/selinux/include/selinux_netlabel.h
@@ -0,0 +1,100 @@
+/*
+ * SELinux interface to the NetLabel subsystem
+ *
+ * Author : Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _SELINUX_NETLABEL_H_
+#define _SELINUX_NETLABEL_H_
+
+#ifdef CONFIG_NETLABEL
+void selinux_netlbl_cache_invalidate(void);
+int selinux_netlbl_inode_permission(struct inode *inode,
+ int mask);
+
+int selinux_netlbl_socket_create(struct socket *sock,
+ const int sock_family,
+ u32 sid);
+int selinux_netlbl_socket_accept(struct socket *sock,
+ struct socket *newsock);
+void selinux_netlbl_socket_post_accept(struct socket *sock,
+ struct socket *newsock);
+int selinux_netlbl_sock_rcv_skb(struct sock *sk,
+ struct sk_buff *skb,
+ const u16 sock_class,
+ const u32 sock_sid,
+ struct avc_audit_data *ad);
+u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock);
+u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb);
+#else
+static inline void selinux_netlbl_cache_invalidate(void)
+{
+ return;
+}
+
+static inline int selinux_netlbl_inode_permission(struct inode *inode,
+ int mask)
+{
+ return 0;
+}
+
+static inline int selinux_netlbl_socket_create(struct socket *sock,
+ const int sock_family,
+ u32 sid)
+{
+ return 0;
+}
+
+static inline int selinux_netlbl_socket_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ return 0;
+}
+
+static inline void selinux_netlbl_socket_post_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ return;
+}
+
+static inline int selinux_netlbl_sock_rcv_skb(struct sock *sk,
+ struct sk_buff *skb,
+ const u16 sock_class,
+ const u32 sock_sid,
+ struct avc_audit_data *ad)
+{
+ return 0;
+}
+
+static inline u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock)
+{
+ return SECSID_NULL;
+}
+
+static inline u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb)
+{
+ return SECSID_NULL;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif
Index: linux-2.6.18-rc1/security/selinux/ss/ebitmap.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/ebitmap.c
+++ linux-2.6.18-rc1/security/selinux/ss/ebitmap.c
@@ -3,6 +3,14 @@
*
* Author : Stephen Smalley, <sds@epoch.ncsc.mil>
*/
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added ebitmap_export() and ebitmap_import()
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
+
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/errno.h>
@@ -59,6 +67,150 @@ int ebitmap_cpy(struct ebitmap *dst, str
return 0;
}
+/**
+ * ebitmap_export - Export an ebitmap to a unsigned char bitmap string
+ * @src: the ebitmap to export
+ * @dst: the resulting bitmap string
+ * @dst_len: length of dst in bytes
+ *
+ * Description:
+ * Allocate a buffer at least src->highbit bits long and export the extensible
+ * bitmap into the buffer. The bitmap string will be in little endian format,
+ * i.e. LSB first. The value returned in dst_len may not the true size of the
+ * buffer as the length of the buffer is rounded up to a multiple of MAPTYPE.
+ * The caller must free the buffer when finished. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int ebitmap_export(const struct ebitmap *src,
+ unsigned char **dst,
+ size_t *dst_len)
+{
+ size_t bitmap_len;
+ unsigned char *bitmap;
+ struct ebitmap_node *iter_node;
+ MAPTYPE node_val;
+ size_t bitmap_byte;
+ unsigned char bitmask;
+
+ if (src == NULL || dst == NULL || dst_len == NULL)
+ return -EINVAL;
+
+ *dst = NULL;
+ *dst_len = 0;
+
+ bitmap_len = src->highbit / 8;
+ if (src->highbit % 8 > 0)
+ bitmap_len += 1;
+ if (bitmap_len == 0)
+ return -EINVAL;
+
+ bitmap = kzalloc(bitmap_len + sizeof(MAPTYPE) -
+ (bitmap_len % sizeof(MAPTYPE)),
+ GFP_ATOMIC);
+ if (bitmap == NULL)
+ return -ENOMEM;
+
+ iter_node = src->node;
+ do {
+ bitmap_byte = iter_node->startbit / 8;
+ bitmask = 0x80;
+ node_val = iter_node->map;
+ do {
+ if (bitmask == 0) {
+ bitmap_byte++;
+ bitmask = 0x80;
+ }
+ if (node_val & (MAPTYPE)0x01)
+ bitmap[bitmap_byte] |= bitmask;
+ node_val >>= 1;
+ bitmask >>= 1;
+ } while (node_val > 0);
+ iter_node = iter_node->next;
+ } while (iter_node);
+
+ *dst = bitmap;
+ *dst_len = bitmap_len;
+ return 0;
+}
+
+/**
+ * ebitmap_import - Import an unsigned char bitmap string into an ebitmap
+ * @src: the bitmap string
+ * @src_len: the bitmap length in bytes
+ * @dst: the empty ebitmap
+ *
+ * Description:
+ * This function takes a little endian bitmap string in src and imports it into
+ * the ebitmap pointed to by dst. Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int ebitmap_import(const unsigned char *src,
+ const size_t src_len,
+ struct ebitmap *dst)
+{
+ size_t src_off = 0;
+ struct ebitmap_node *node_new;
+ struct ebitmap_node *node_last = NULL;
+ size_t src_rem = src_len;
+ MAPTYPE tmp_val;
+ size_t iter;
+ size_t iter_bit;
+ unsigned char src_byte;
+
+ if (src == NULL || dst == NULL || src_len == 0)
+ return -EINVAL;
+
+ do {
+ node_new = kzalloc(sizeof(*node_new), GFP_ATOMIC);
+ if (node_new == NULL) {
+ ebitmap_destroy(dst);
+ return -ENOMEM;
+ }
+
+ if (src_rem >= sizeof(MAPTYPE))
+ iter = sizeof(MAPTYPE);
+ else
+ iter = src_rem;
+ tmp_val = 0;
+ while (iter > 0) {
+ src_byte = src[src_off + --iter];
+ if (src_byte > 0)
+ for (iter_bit = 0; iter_bit < 8; iter_bit++) {
+ tmp_val <<= 1;
+ tmp_val |= src_byte & 0x01;
+ src_byte >>= 1;
+ }
+ else
+ tmp_val <<= 8;
+ }
+ node_new->map = tmp_val;
+ node_new->startbit = src_off * 8;
+
+ if (node_last != NULL)
+ node_last->next = node_new;
+ else
+ dst->node = node_new;
+ node_last = node_new;
+
+ if (src_rem >= sizeof(MAPTYPE)) {
+ src_off += sizeof(MAPTYPE);
+ src_rem -= sizeof(MAPTYPE);
+ } else
+ src_off += src_rem;
+ } while (src_off < src_len);
+
+ tmp_val = node_last->map;
+ dst->highbit = node_last->startbit;
+ while (tmp_val >= 1) {
+ dst->highbit += 1;
+ tmp_val >>= 1;
+ }
+
+ return 0;
+}
+
int ebitmap_contains(struct ebitmap *e1, struct ebitmap *e2)
{
struct ebitmap_node *n1, *n2;
Index: linux-2.6.18-rc1/security/selinux/ss/ebitmap.h
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/ebitmap.h
+++ linux-2.6.18-rc1/security/selinux/ss/ebitmap.h
@@ -69,6 +69,12 @@ static inline int ebitmap_node_get_bit(s
int ebitmap_cmp(struct ebitmap *e1, struct ebitmap *e2);
int ebitmap_cpy(struct ebitmap *dst, struct ebitmap *src);
+int ebitmap_export(const struct ebitmap *src,
+ unsigned char **dst,
+ size_t *dst_len);
+int ebitmap_import(const unsigned char *src,
+ const size_t src_len,
+ struct ebitmap *dst);
int ebitmap_contains(struct ebitmap *e1, struct ebitmap *e2);
int ebitmap_get_bit(struct ebitmap *e, unsigned long bit);
int ebitmap_set_bit(struct ebitmap *e, unsigned long bit, int value);
Index: linux-2.6.18-rc1/security/selinux/ss/mls.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/mls.c
+++ linux-2.6.18-rc1/security/selinux/ss/mls.c
@@ -10,6 +10,13 @@
*
* Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
*/
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added support to import/export the MLS label
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
#include <linux/kernel.h>
#include <linux/slab.h>
@@ -585,3 +592,148 @@ int mls_compute_sid(struct context *scon
return -EINVAL;
}
+/**
+ * mls_export_lvl - Export the MLS sensitivity levels
+ * @context: the security context
+ * @low: the low sensitivity level
+ * @high: the high sensitivity level
+ *
+ * Description:
+ * Given the security context copy the low MLS sensitivity level into lvl_low
+ * and the high sensitivity level in lvl_high. The MLS levels are only
+ * exported if the pointers are not NULL, if they are NULL then that level is
+ * not exported.
+ *
+ */
+void mls_export_lvl(const struct context *context, u32 *low, u32 *high)
+{
+ if (!selinux_mls_enabled)
+ return;
+
+ if (low != NULL)
+ *low = context->range.level[0].sens - 1;
+ if (high != NULL)
+ *high = context->range.level[1].sens - 1;
+}
+
+/**
+ * mls_import_lvl - Import the MLS sensitivity levels
+ * @context: the security context
+ * @low: the low sensitivity level
+ * @high: the high sensitivity level
+ *
+ * Description:
+ * Given the security context and the two sensitivty levels, set the MLS levels
+ * in the context according the two given as parameters. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+void mls_import_lvl(struct context *context, const u32 low, const u32 high)
+{
+ if (!selinux_mls_enabled)
+ return;
+
+ context->range.level[0].sens = low + 1;
+ context->range.level[1].sens = high + 1;
+}
+
+/**
+ * mls_export_cat - Export the MLS categories
+ * @context: the security context
+ * @low: the low category
+ * @low_len: length of the cat_low bitmap in bytes
+ * @high: the high category
+ * @high_len: length of the cat_high bitmap in bytes
+ *
+ * Description:
+ * Given the security context export the low MLS category bitmap into cat_low
+ * and the high category bitmap into cat_high. The MLS categories are only
+ * exported if the pointers are not NULL, if they are NULL then that level is
+ * not exported. The caller is responsibile for freeing the memory when
+ * finished. Returns zero on success, negative values on failure.
+ *
+ */
+int mls_export_cat(const struct context *context,
+ unsigned char **low,
+ size_t *low_len,
+ unsigned char **high,
+ size_t *high_len)
+{
+ int rc = -EPERM;
+
+ if (!selinux_mls_enabled)
+ return 0;
+
+ if (low != NULL) {
+ rc = ebitmap_export(&context->range.level[0].cat,
+ low,
+ low_len);
+ if (rc != 0)
+ goto export_cat_failure;
+ }
+ if (high != NULL) {
+ rc = ebitmap_export(&context->range.level[1].cat,
+ high,
+ high_len);
+ if (rc != 0)
+ goto export_cat_failure;
+ }
+
+ return 0;
+
+export_cat_failure:
+ if (low != NULL)
+ kfree(*low);
+ if (high != NULL)
+ kfree(*high);
+ return rc;
+}
+
+/**
+ * mls_import_cat - Import the MLS categories
+ * @context: the security context
+ * @low: the low category
+ * @low_len: length of the cat_low bitmap in bytes
+ * @high: the high category
+ * @high_len: length of the cat_high bitmap in bytes
+ *
+ * Description:
+ * Given the security context and the two category bitmap strings import the
+ * categories into the security context. The MLS categories are only imported
+ * if the pointers are not NULL, if they are NULL they are skipped. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+int mls_import_cat(struct context *context,
+ const unsigned char *low,
+ const size_t low_len,
+ const unsigned char *high,
+ const size_t high_len)
+{
+ int rc = -EPERM;
+
+ if (!selinux_mls_enabled)
+ return 0;
+
+ if (low != NULL) {
+ rc = ebitmap_import(low,
+ low_len,
+ &context->range.level[0].cat);
+ if (rc != 0)
+ goto import_cat_failure;
+ }
+ if (high != NULL) {
+ rc = ebitmap_import(high,
+ high_len,
+ &context->range.level[1].cat);
+ if (rc != 0)
+ goto import_cat_failure;
+ }
+
+ return 0;
+
+import_cat_failure:
+ ebitmap_destroy(&context->range.level[0].cat);
+ ebitmap_destroy(&context->range.level[1].cat);
+ return rc;
+}
Index: linux-2.6.18-rc1/security/selinux/ss/mls.h
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/mls.h
+++ linux-2.6.18-rc1/security/selinux/ss/mls.h
@@ -10,6 +10,13 @@
*
* Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
*/
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added support to import/export the MLS label
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
#ifndef _SS_MLS_H_
#define _SS_MLS_H_
@@ -42,5 +49,19 @@ int mls_compute_sid(struct context *scon
int mls_setup_user_range(struct context *fromcon, struct user_datum *user,
struct context *usercon);
+void mls_export_lvl(const struct context *context, u32 *low, u32 *high);
+void mls_import_lvl(struct context *context, const u32 low, const u32 high);
+
+int mls_export_cat(const struct context *context,
+ unsigned char **low,
+ size_t *low_len,
+ unsigned char **high,
+ size_t *high_len);
+int mls_import_cat(struct context *context,
+ const unsigned char *low,
+ const size_t low_len,
+ const unsigned char *high,
+ const size_t high_len);
+
#endif /* _SS_MLS_H */
Index: linux-2.6.18-rc1/security/selinux/ss/services.c
===================================================================
--- linux-2.6.18-rc1.orig/security/selinux/ss/services.c
+++ linux-2.6.18-rc1/security/selinux/ss/services.c
@@ -13,6 +13,11 @@
*
* Added conditional policy language extensions
*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ * Added support for NetLabel
+ *
+ * Copyright (C) 2006 Hewlett-Packard Development Company, L.P.
* Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
* Copyright (C) 2003 - 2004 Tresys Technology, LLC
* Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
@@ -29,6 +34,8 @@
#include <linux/sched.h>
#include <linux/audit.h>
#include <linux/mutex.h>
+#include <net/sock.h>
+#include <net/netlabel.h>
#include "flask.h"
#include "avc.h"
@@ -40,6 +47,8 @@
#include "services.h"
#include "conditional.h"
#include "mls.h"
+#include "objsec.h"
+#include "selinux_netlabel.h"
extern void selnl_notify_policyload(u32 seqno);
unsigned int policydb_loaded_version;
@@ -1241,6 +1250,7 @@ int security_load_policy(void *data, siz
selinux_complete_init();
avc_ss_reset(seqno);
selnl_notify_policyload(seqno);
+ selinux_netlbl_cache_invalidate();
return 0;
}
@@ -1295,6 +1305,7 @@ int security_load_policy(void *data, siz
avc_ss_reset(seqno);
selnl_notify_policyload(seqno);
+ selinux_netlbl_cache_invalidate();
return 0;
@@ -2064,3 +2075,547 @@ void selinux_audit_set_callback(int (*ca
{
aurule_callback = callback;
}
+
+#ifdef CONFIG_NETLABEL
+/*
+ * This is the structure we store inside the NetLabel cache block.
+ */
+#define NETLBL_CACHE(x) ((struct netlbl_cache *)(x))
+#define NETLBL_CACHE_T_NONE 0
+#define NETLBL_CACHE_T_SID 1
+#define NETLBL_CACHE_T_MLSLBL 2
+struct netlbl_cache {
+ u32 type;
+ union {
+ u32 sid;
+ struct mls_level mls_label;
+ } data;
+};
+
+/**
+ * selinux_netlbl_cache_free - Free the NetLabel cached data
+ * @data: the data to free
+ *
+ * Description:
+ * This function is intended to be used as the free() callback inside the
+ * netlbl_lsm_cache structure.
+ *
+ */
+static void selinux_netlbl_cache_free(const void *data)
+{
+ struct netlbl_cache *cache = NETLBL_CACHE(data);
+ switch (cache->type) {
+ case NETLBL_CACHE_T_MLSLBL:
+ ebitmap_destroy(&cache->data.mls_label.cat);
+ break;
+ }
+ kfree(data);
+}
+
+/**
+ * selinux_netlbl_cache_add - Add an entry to the NetLabel cache
+ * @skb: the packet
+ * @ctx: the SELinux context
+ *
+ * Description:
+ * Attempt to cache the context in @ctx, which was derived from the packet in
+ * @skb, in the NetLabel subsystem cache.
+ *
+ */
+static void selinux_netlbl_cache_add(struct sk_buff *skb, struct context *ctx)
+{
+ struct netlbl_cache *cache = NULL;
+ struct netlbl_lsm_secattr secattr;
+ u32 cache_clear = 1;
+
+ netlbl_secattr_init(&secattr);
+
+ cache = kzalloc(sizeof(*cache), GFP_ATOMIC);
+ if (cache == NULL)
+ goto netlbl_cache_add_return;
+ secattr.cache.free = selinux_netlbl_cache_free;
+ secattr.cache.data = (void *)cache;
+
+ cache->type = NETLBL_CACHE_T_MLSLBL;
+ if (ebitmap_cpy(&cache->data.mls_label.cat,
+ &ctx->range.level[0].cat) != 0)
+ goto netlbl_cache_add_return;
+ cache->data.mls_label.sens = ctx->range.level[0].sens;
+
+ if (netlbl_cache_add(skb, &secattr) != 0)
+ goto netlbl_cache_add_return;
+ cache_clear = 0;
+
+netlbl_cache_add_return:
+ netlbl_secattr_destroy(&secattr, cache_clear);
+}
+
+/**
+ * selinux_netlbl_cache_invalidate - Invalidate the NetLabel cache
+ *
+ * Description:
+ * Invalidate the NetLabel security attribute mapping cache.
+ *
+ */
+void selinux_netlbl_cache_invalidate(void)
+{
+ netlbl_cache_invalidate();
+}
+
+/**
+ * selinux_netlbl_secattr_to_sid - Convert a NetLabel secattr to a SELinux SID
+ * @skb: the network packet
+ * @secattr: the NetLabel packet security attributes
+ * @base_sid: the SELinux SID to use as a context for MLS only attributes
+ * @sid: the SELinux SID
+ *
+ * Description:
+ * Convert the given NetLabel packet security attributes in @secattr into a
+ * SELinux SID. If the @secattr field does not contain a full SELinux
+ * SID/context then use the context in @base_sid as the foundation. If @skb
+ * is not NULL attempt to cache as much data as possibile. Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_secattr_to_sid(struct sk_buff *skb,
+ struct netlbl_lsm_secattr *secattr,
+ const u32 base_sid,
+ u32 *sid)
+{
+ int rc = -EIDRM;
+ struct context *ctx;
+ struct context ctx_new;
+ struct netlbl_cache *cache;
+ u32 ctx_new_destroy = 0;
+
+ POLICY_RDLOCK;
+
+ if (secattr->cache.data) {
+ cache = NETLBL_CACHE(secattr->cache.data);
+ switch (cache->type) {
+ case NETLBL_CACHE_T_SID:
+ *sid = cache->data.sid;
+ break;
+ case NETLBL_CACHE_T_MLSLBL:
+ ctx = sidtab_search(&sidtab, base_sid);
+ if (ctx == NULL)
+ goto netlbl_secattr_to_sid_return;
+ rc = context_cpy(&ctx_new, ctx);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+ ctx_new_destroy = 1;
+ mls_context_destroy(&ctx_new);
+
+ ctx_new.range.level[0].sens =
+ cache->data.mls_label.sens;
+ rc = ebitmap_cpy(&ctx_new.range.level[0].cat,
+ &cache->data.mls_label.cat);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+ ctx_new.range.level[1].sens =
+ cache->data.mls_label.sens;
+ rc = ebitmap_cpy(&ctx_new.range.level[1].cat,
+ &cache->data.mls_label.cat);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+
+ rc = sidtab_context_to_sid(&sidtab, &ctx_new, sid);
+ break;
+ default:
+ goto netlbl_secattr_to_sid_return;
+ }
+ } else if (secattr->mls_lvl_vld) {
+ ctx = sidtab_search(&sidtab, base_sid);
+ if (ctx == NULL)
+ goto netlbl_secattr_to_sid_return;
+ rc = context_cpy(&ctx_new, ctx);
+ if (rc != 0)
+ goto netlbl_secattr_to_sid_return;
+ ctx_new_destroy = 1;
+ mls_context_destroy(&ctx_new);
+
+ mls_import_lvl(&ctx_new, secattr->mls_lvl, secattr->mls_lvl);
+ if (secattr->mls_cat) {
+ if (mls_import_cat(&ctx_new,
+ secattr->mls_cat,
+ secattr->mls_cat_len,
+ secattr->mls_cat,
+ secattr->mls_cat_len) != 0)
+ goto netlbl_secattr_to_sid_return;
+ }
+
+ rc = mls_context_isvalid(&policydb, &ctx_new);
+ if (rc != 1)
+ goto netlbl_secattr_to_sid_return;
+
+ if (skb != NULL)
+ selinux_netlbl_cache_add(skb, &ctx_new);
+
+ rc = sidtab_context_to_sid(&sidtab, &ctx_new, sid);
+ } else
+ *sid = SECINITSID_UNLABELED;
+
+ rc = 0;
+
+netlbl_secattr_to_sid_return:
+ POLICY_RDUNLOCK;
+ if (ctx_new_destroy)
+ context_destroy(&ctx_new);
+ return rc;
+}
+
+/**
+ * selinux_netlbl_skbuff_getsid - Get the sid of a packet using NetLabel
+ * @skb: the packet
+ * @base_sid: the SELinux SID to use as a context for MLS only attributes
+ * @sid: the SID
+ *
+ * Description:
+ * Call the NetLabel mechanism to get the security attributes of the given
+ * packet and use those attributes to determine the correct context/SID to
+ * assign to the packet. Returns zero on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_skbuff_getsid(struct sk_buff *skb,
+ const u32 base_sid,
+ u32 *sid)
+{
+ int rc;
+ struct netlbl_lsm_secattr secattr;
+
+ netlbl_secattr_init(&secattr);
+ rc = netlbl_skbuff_getattr(skb, &secattr);
+ if (rc == 0)
+ rc = selinux_netlbl_secattr_to_sid(skb,
+ &secattr,
+ base_sid,
+ sid);
+ netlbl_secattr_destroy(&secattr, 0);
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_setsid - Label a socket using the NetLabel mechanism
+ * @sock: the socket to label
+ * @sock_family: the socket family
+ * @sid: the SID to use
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism using the given
+ * SID. Returns zero values on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_socket_setsid(struct socket *sock,
+ const int sock_family,
+ u32 sid)
+{
+ int rc;
+ struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+ struct netlbl_lsm_secattr secattr;
+ struct context *ctx;
+
+ if (!ss_initialized)
+ return 0;
+
+ POLICY_RDLOCK;
+
+ ctx = sidtab_search(&sidtab, sid);
+ if (ctx != NULL) {
+ netlbl_secattr_init(&secattr);
+ secattr.domain = kstrdup(
+ policydb.p_type_val_to_name[ctx->type - 1],
+ GFP_ATOMIC);
+ mls_export_lvl(ctx, &secattr.mls_lvl, NULL);
+ secattr.mls_lvl_vld = 1;
+ mls_export_cat(ctx,
+ &secattr.mls_cat,
+ &secattr.mls_cat_len,
+ NULL,
+ NULL);
+
+ rc = netlbl_socket_setattr(sock, &secattr);
+ if (rc == 0)
+ isec->netlbl.state = NLBL_LABELED;
+
+ netlbl_secattr_destroy(&secattr, 0);
+ } else
+ rc = -ENOENT;
+
+ POLICY_RDUNLOCK;
+
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_peeksid - Get the SID of the first queued packet
+ * @sock: the socket to query
+ * @sid: the packet's SID
+ *
+ * Description:
+ * Examine the first incoming packet in the socket's queue and determine the
+ * packet's SELinux SID. Return zero on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_socket_peeksid(struct socket *sock,
+ u32 *sid)
+{
+ int rc;
+ struct netlbl_lsm_secattr secattr;
+ struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+
+ netlbl_secattr_init(&secattr);
+ rc = netlbl_socket_peekattr(sock, &secattr);
+ if (rc == 0)
+ rc = selinux_netlbl_secattr_to_sid(NULL,
+ &secattr,
+ isec->sid,
+ sid);
+ netlbl_secattr_destroy(&secattr, 0);
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_create - Label a socket using the NetLabel mechanism
+ * @sock: the socket to label
+ * @sock_family: the socket family
+ * @sid: the SID to use
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism using the given
+ * SID. Returns zero values on success, negative values on failure.
+ *
+ */
+int selinux_netlbl_socket_create(struct socket *sock,
+ const int sock_family,
+ u32 sid)
+{
+ struct inode_security_struct *isec;
+
+ if (sock_family != PF_INET)
+ return 0;
+
+ isec = SOCK_INODE(sock)->i_security;
+ isec->netlbl.state = NLBL_REQUIRE;
+ return selinux_netlbl_socket_setsid(sock, sock_family, sid);
+}
+
+/**
+ * selinux_netlbl_socket_accept - NetLabel accept hook
+ * @sock: the original socket
+ * @newsock: the new accept()ed socket
+ *
+ * Description:
+ * Mark @newsock as requiring a NetLabel label and do setup required for
+ * further processing by the post_accept and sock_rcv_skb hooks. Returns zero
+ * on success, negative values on failure.
+ *
+ */
+int selinux_netlbl_socket_accept(struct socket *sock, struct socket *newsock)
+{
+ struct inode_security_struct *isec;
+ struct inode_security_struct *newisec;
+
+ isec = SOCK_INODE(sock)->i_security;
+ if (isec->netlbl.state == NLBL_UNSET)
+ return 0;
+
+ newisec = SOCK_INODE(newsock)->i_security;
+ newisec->netlbl.state = NLBL_ACCEPT;
+
+ return 0;
+}
+
+/**
+ * selinux_netlbl_socket_post_accept - NetLabel post_accept hook
+ * @sock: the original socket
+ * @newsock: the new accept()ed socket
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism based on the
+ * connection and the original/parent socket's SID.
+ *
+ */
+void selinux_netlbl_socket_post_accept(struct socket *sock,
+ struct socket *newsock)
+{
+ struct inode_security_struct *newisec;
+ struct netlbl_lsm_secattr secattr;
+
+ newisec = SOCK_INODE(newsock)->i_security;
+ if (newisec->netlbl.state != NLBL_ACCEPT)
+ return;
+ netlbl_secattr_init(&secattr);
+ if (netlbl_socket_getattr(newsock, &secattr) != 0)
+ goto netlbl_socket_post_accept_return;
+ newisec = SOCK_INODE(newsock)->i_security;
+ if (selinux_netlbl_secattr_to_sid(NULL,
+ &secattr,
+ newisec->sid,
+ &newisec->netlbl.peer_sid) != 0)
+ goto netlbl_socket_post_accept_return;
+ newisec->sid = newisec->netlbl.peer_sid;
+
+ /* XXX - try to set the NetLabel here to try and save some time in
+ selinux_netlbl_inode_permission() later */
+ selinux_netlbl_socket_setsid(newsock,
+ newsock->sk->sk_family,
+ newisec->sid);
+
+netlbl_socket_post_accept_return:
+ if (newisec->netlbl.state != NLBL_LABELED)
+ newisec->netlbl.state = NLBL_REQUIRE;
+ netlbl_secattr_destroy(&secattr, 0);
+}
+
+/**
+ * selinux_netlbl_inode_permission - Verify the file/socket is NetLabel labeled
+ * @inode: the file descriptor's inode
+ * @mask: the permission mask
+ *
+ * Description:
+ * Looks at a file's inode and if it is marked as a socket protected by
+ * NetLabel then verify that the socket has been labeled, if not try to label
+ * the socket now with the task's SID. Returns zero on success, negative
+ * values on failure.
+ *
+ */
+int selinux_netlbl_inode_permission(struct inode *inode, int mask)
+{
+ int rc = 0;
+ struct inode_security_struct *isec = inode->i_security;
+ struct netlbl_security_struct *nlblsec = &isec->netlbl;
+ struct socket *sock;
+
+ down(&isec->sem);
+ if (nlblsec->state == NLBL_REQUIRE &&
+ (mask & (MAY_WRITE | MAY_APPEND))) {
+ sock = SOCKET_I(inode);
+ rc = selinux_netlbl_socket_setsid(sock,
+ sock->sk->sk_family,
+ isec->sid);
+ }
+ up(&isec->sem);
+
+ return rc;
+}
+
+/**
+ * selinux_netlbl_sock_rcv_skb - Do an inbound access check using NetLabel
+ * @sk: the packet's sock
+ * @skb: the packet
+ * @sock_class: the socket class
+ * @sock_sid: the socket's SID
+ * @ad: the audit data
+ *
+ * Description:
+ * Fetch the NetLabel security attributes from @skb and perform an access check
+ * against the receiving socket. Returns zero on success, negative values on
+ * error.
+ *
+ */
+int selinux_netlbl_sock_rcv_skb(struct sock *sk,
+ struct sk_buff *skb,
+ const u16 sock_class,
+ const u32 sock_sid,
+ struct avc_audit_data *ad)
+{
+ int rc;
+ u32 netlbl_sid;
+ u32 recv_perm;
+
+ rc = selinux_netlbl_skbuff_getsid(skb, sock_sid, &netlbl_sid);
+ if (rc != 0)
+ return rc;
+
+ if (netlbl_sid != SECINITSID_UNLABELED) {
+ switch (sock_class) {
+ case SECCLASS_UDP_SOCKET:
+ recv_perm = UDP_SOCKET__RECV_MSG;
+ break;
+ case SECCLASS_TCP_SOCKET:
+ recv_perm = TCP_SOCKET__RECV_MSG;
+ break;
+ default:
+ recv_perm = RAWIP_SOCKET__RECV_MSG;
+ break;
+ }
+
+ rc = avc_has_perm(sock_sid,
+ netlbl_sid,
+ sock_class,
+ recv_perm,
+ ad);
+ if (rc != 0) {
+ struct inode_security_struct *isec;
+ u32 netlbl_send_err = 1;
+
+ read_lock_bh(&sk->sk_callback_lock);
+ isec = SOCK_INODE(sk->sk_socket)->i_security;
+ if (isec->netlbl.state == NLBL_ACCEPT)
+ netlbl_send_err = 0;
+ read_unlock_bh(&sk->sk_callback_lock);
+
+ if (netlbl_send_err)
+ netlbl_skbuff_err(skb, rc);
+ }
+ }
+
+ return rc;
+}
+
+/**
+ * selinux_netlbl_socket_peersid - Return the peer SID of a connected socket
+ * @sock: the socket
+ *
+ * Description:
+ * Examine @sock to find the connected peer's SID, if that is not possibile
+ * check the socket's receive queue and take the SID from the incoming
+ * packets. Returns the SID on success or SECSID_NULL on error.
+ *
+ */
+u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock)
+{
+ u32 peer_sid;
+ struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+
+ if (isec->netlbl.peer_sid != 0)
+ peer_sid = isec->netlbl.peer_sid;
+ else {
+ if (selinux_netlbl_socket_peeksid(sock, &peer_sid) != 0)
+ peer_sid = SECSID_NULL;
+ }
+
+ return peer_sid;
+}
+
+/**
+ * selinux_netlbl_socket_getpeersec_dgram - Return the SID of a NetLabel packet
+ * @skb: the packet
+ *
+ * Description:
+ * Examine @skb to find the SID assigned to it by NetLabel. Returns the SID on
+ * success, SECSID_NULL on error.
+ *
+ */
+u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb)
+{
+ int peer_sid;
+ struct sock *sk = skb->sk;
+ struct inode_security_struct *isec;
+
+ if (sk == NULL)
+ return SECSID_NULL;
+
+ read_lock_bh(&skb->sk->sk_callback_lock);
+ if (skb->sk->sk_socket == NULL) {
+ read_unlock_bh(&skb->sk->sk_callback_lock);
+ return SECSID_NULL;
+ }
+ isec = SOCK_INODE(skb->sk->sk_socket)->i_security;
+ if (selinux_netlbl_skbuff_getsid(skb, isec->sid, &peer_sid) != 0)
+ peer_sid = SECSID_NULL;
+ read_unlock_bh(&sk->sk_callback_lock);
+
+ return peer_sid;
+}
+#endif /* CONFIG_NETLABEL */
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 7/7] NetLabel: tie NetLabel into the Kconfig system
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 18:57 ` paul.moore
-1 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-net_kconfig-2.6.18 --]
[-- Type: text/plain, Size: 553 bytes --]
Modify the net/Kconfig file to enable selecting the NetLabel Kconfig options.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/Kconfig | 2 ++
1 files changed, 2 insertions(+)
Index: linux-2.6.18-rc1/net/Kconfig
===================================================================
--- linux-2.6.18-rc1.orig/net/Kconfig
+++ linux-2.6.18-rc1/net/Kconfig
@@ -249,6 +249,8 @@ source "net/ieee80211/Kconfig"
config WIRELESS_EXT
bool
+source "net/netlabel/Kconfig"
+
endif # if NET
endmenu # Networking
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 7/7] NetLabel: tie NetLabel into the Kconfig system
@ 2006-07-14 18:57 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-14 18:57 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Modify the net/Kconfig file to enable selecting the NetLabel Kconfig options.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/Kconfig | 2 ++
1 files changed, 2 insertions(+)
Index: linux-2.6.18-rc1/net/Kconfig
===================================================================
--- linux-2.6.18-rc1.orig/net/Kconfig
+++ linux-2.6.18-rc1/net/Kconfig
@@ -249,6 +249,8 @@ source "net/ieee80211/Kconfig"
config WIRELESS_EXT
bool
+source "net/netlabel/Kconfig"
+
endif # if NET
endmenu # Networking
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 2/7] NetLabel: core network changes
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 23:34 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-14 23:34 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +static inline void netlbl_put_hdr(unsigned char *buffer,
> + const u32 msg_type,
> + const u16 msg_len,
> + const u16 msg_flags,
> + const u32 msg_pid,
> + const u32 msg_seq)
Why are these parameters marked const? Seems to be common throughout the
code.
- James
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 2/7] NetLabel: core network changes
@ 2006-07-14 23:34 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-14 23:34 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +static inline void netlbl_put_hdr(unsigned char *buffer,
> + const u32 msg_type,
> + const u16 msg_len,
> + const u16 msg_flags,
> + const u32 msg_pid,
> + const u32 msg_seq)
Why are these parameters marked const? Seems to be common throughout the
code.
- James
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 2/7] NetLabel: core network changes
2006-07-14 23:34 ` James Morris
(?)
@ 2006-07-14 23:36 ` David Miller
-1 siblings, 0 replies; 55+ messages in thread
From: David Miller @ 2006-07-14 23:36 UTC (permalink / raw)
To: jmorris; +Cc: paul.moore, netdev, selinux, sds, pratt
From: James Morris <jmorris@namei.org>
Date: Fri, 14 Jul 2006 19:34:20 -0400 (EDT)
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
>
> > +static inline void netlbl_put_hdr(unsigned char *buffer,
> > + const u32 msg_type,
> > + const u16 msg_len,
> > + const u16 msg_flags,
> > + const u32 msg_pid,
> > + const u32 msg_seq)
>
> Why are these parameters marked const? Seems to be common throughout the
> code.
Yes, it is pointless to mark non-pointers as "const", especially
these day.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-14 23:56 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-14 23:56 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +int cipso_v4_validate(unsigned char **option)
> +{
> + unsigned char *opt = *option;
> + unsigned char *tag;
> + unsigned char opt_iter;
> + unsigned char err_offset = 0;
> + unsigned char locked = 0;
> + u8 opt_len;
> + u8 tag_len;
> + struct cipso_v4_doi *doi_def = NULL;
> + u32 tag_iter;
> +
> + /* XXX - caller already checks for length values that are too large */
Please don't put 'XXX' there for a normaly comment. It's supposed to
indicate that something needs attention (e.g. known bug, unresolved
question etc).
> + rcu_read_lock();
> + locked = 1;
> +validate_return:
> + if (locked)
> + rcu_read_unlock();
> + *option = opt + err_offset;
> + return err_offset;
> +}
No need for the 'locked' variable, just create a new goto label, like:
{
out_unlock:
rcu_read_unlock();
out:
*option = opt + err_offset;
return err_offset;
}
Then have the goto callers specify the correct label.
> + /* In the case of sock_create_lite(), the sock->sk field is not
> + defined yet but it is not a problem as the only users of these
> + "lite" PF_INET sockets are functions which do an accept() call
> + afterwards so we will label the socket as part of the accept(). */
Coding style standard for multiline comments is:
/*
*
*/
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-14 23:56 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-14 23:56 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +int cipso_v4_validate(unsigned char **option)
> +{
> + unsigned char *opt = *option;
> + unsigned char *tag;
> + unsigned char opt_iter;
> + unsigned char err_offset = 0;
> + unsigned char locked = 0;
> + u8 opt_len;
> + u8 tag_len;
> + struct cipso_v4_doi *doi_def = NULL;
> + u32 tag_iter;
> +
> + /* XXX - caller already checks for length values that are too large */
Please don't put 'XXX' there for a normaly comment. It's supposed to
indicate that something needs attention (e.g. known bug, unresolved
question etc).
> + rcu_read_lock();
> + locked = 1;
> +validate_return:
> + if (locked)
> + rcu_read_unlock();
> + *option = opt + err_offset;
> + return err_offset;
> +}
No need for the 'locked' variable, just create a new goto label, like:
{
out_unlock:
rcu_read_unlock();
out:
*option = opt + err_offset;
return err_offset;
}
Then have the goto callers specify the correct label.
> + /* In the case of sock_create_lite(), the sock->sk field is not
> + defined yet but it is not a problem as the only users of these
> + "lite" PF_INET sockets are functions which do an accept() call
> + afterwards so we will label the socket as part of the accept(). */
Coding style standard for multiline comments is:
/*
*
*/
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-15 2:03 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 2:03 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +/**
> + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
Can you use lib/bitmap.c instead?
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 2:03 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 2:03 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +/**
> + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
Can you use lib/bitmap.c instead?
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-15 2:18 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 2:18 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +
> +/* Label mapping cache */
> +int cipso_v4_cache_enabled = 1;
> +int cipso_v4_cache_bucketsize = 10;
> +static struct cipso_v4_map_cache_bkt *cipso_v4_cache = NULL;
> +#define CIPSO_V4_CACHE_ENABLED (cipso_v4_cache_enabled && \
> + cipso_v4_cache_bucketsize > 0 && \
> + cipso_v4_cache_size > 0)
This macro seems kind of odd. Why not just check for
cipso_v4_cache_enabled ?
> +static int cipso_v4_cache_init(const u32 bkt_size)
> +{
> + struct cipso_v4_map_cache_bkt *cache;
Why have this temporary local variable,
> + cipso_v4_cache = cache;
then assign its value to a global?
This function is only ever called once, right?
> + * Description:
> + * Clears the CIPSO cache and frees all the memory. This function does not
> + * hold any locks and should only be called when the module is being unloaded.
> + *
> + */
> +static int cipso_v4_cache_destroy(void)
> +{
> + struct cipso_v4_map_cache_bkt *cache;
> + struct cipso_v4_map_cache_entry *entry;
> + u32 cache_size;
> + u32 iter;
> +
> + cache = cipso_v4_cache;
> + cache_size = cipso_v4_cache_size;
> + cipso_v4_cache_size = 0;
> + cipso_v4_cache = NULL;
> +
> + for (iter = 0; iter < cache_size; iter++)
> + list_for_each_entry(entry, &cache[iter].list, list) {
> + list_del(&entry->list);
> + cipso_v4_cache_entry_free(entry);
> + }
> +
> + return 0;
> +}
> +
If you're in a permanent cleanup phase, why bother clearing all of these
global variables and using temporary pointers?
Why not just simply free each entry? The way this is coded makes it seem
like you're unsure about the safety of the code.
e.g. what could it possibly matter at this stage whether cipso_v4_cache is
NULL or not?
I think I asked this some time ago: what are the lifetime rules for this
code as a loadable module? What if you arbitrarily rmmod it?
- James
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 2:18 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 2:18 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +
> +/* Label mapping cache */
> +int cipso_v4_cache_enabled = 1;
> +int cipso_v4_cache_bucketsize = 10;
> +static struct cipso_v4_map_cache_bkt *cipso_v4_cache = NULL;
> +#define CIPSO_V4_CACHE_ENABLED (cipso_v4_cache_enabled && \
> + cipso_v4_cache_bucketsize > 0 && \
> + cipso_v4_cache_size > 0)
This macro seems kind of odd. Why not just check for
cipso_v4_cache_enabled ?
> +static int cipso_v4_cache_init(const u32 bkt_size)
> +{
> + struct cipso_v4_map_cache_bkt *cache;
Why have this temporary local variable,
> + cipso_v4_cache = cache;
then assign its value to a global?
This function is only ever called once, right?
> + * Description:
> + * Clears the CIPSO cache and frees all the memory. This function does not
> + * hold any locks and should only be called when the module is being unloaded.
> + *
> + */
> +static int cipso_v4_cache_destroy(void)
> +{
> + struct cipso_v4_map_cache_bkt *cache;
> + struct cipso_v4_map_cache_entry *entry;
> + u32 cache_size;
> + u32 iter;
> +
> + cache = cipso_v4_cache;
> + cache_size = cipso_v4_cache_size;
> + cipso_v4_cache_size = 0;
> + cipso_v4_cache = NULL;
> +
> + for (iter = 0; iter < cache_size; iter++)
> + list_for_each_entry(entry, &cache[iter].list, list) {
> + list_del(&entry->list);
> + cipso_v4_cache_entry_free(entry);
> + }
> +
> + return 0;
> +}
> +
If you're in a permanent cleanup phase, why bother clearing all of these
global variables and using temporary pointers?
Why not just simply free each entry? The way this is coded makes it seem
like you're unsure about the safety of the code.
e.g. what could it possibly matter at this stage whether cipso_v4_cache is
NULL or not?
I think I asked this some time ago: what are the lifetime rules for this
code as a loadable module? What if you arbitrarily rmmod it?
- James
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-15 13:08 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 13:08 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
> +{
> + struct sk_buff *skb;
> + unsigned char *buf;
> + struct cipso_v4_doi *iter;
> + u32 doi_cnt = 0;
> + ssize_t buf_len;
> +
> + /* XXX - In both cases, this is kinda ugly as we have to go through
> + the list once to determine how large of a buffer we need,
> + drop the locks, allocate the buffer, grab the locks, and
> + finally fill the buffer. The problem is that there is that
> + open window where the table could grow and we will end up
> + short on space. */
This needs to be fixed.
Currently, it seems that you just bail if this happens.
> +
> + buf_len = NETLBL_LEN_U32;
> + rcu_read_lock();
> + list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
> + if (iter->valid) {
> + doi_cnt += 1;
> + buf_len += 2 * NETLBL_LEN_U32;
> + }
> + rcu_read_unlock();
> +
> + skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
> + if (skb == NULL)
> + return NULL;
> + buf = skb_put(skb, buf_len);
> + netlbl_putinc_u32(&buf, doi_cnt, &buf_len);
> +
> + rcu_read_lock();
> + list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
> + if (iter->valid) {
> + if (buf_len < 2 * NETLBL_LEN_U32) {
> + rcu_read_unlock();
> + kfree(skb);
> + return NULL;
> + }
> + netlbl_putinc_u32(&buf, iter->doi, &buf_len);
> + netlbl_putinc_u32(&buf, iter->type, &buf_len);
> + }
> + rcu_read_unlock();
> +
> + return skb;
> +}
> +
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 13:08 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 13:08 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
> +{
> + struct sk_buff *skb;
> + unsigned char *buf;
> + struct cipso_v4_doi *iter;
> + u32 doi_cnt = 0;
> + ssize_t buf_len;
> +
> + /* XXX - In both cases, this is kinda ugly as we have to go through
> + the list once to determine how large of a buffer we need,
> + drop the locks, allocate the buffer, grab the locks, and
> + finally fill the buffer. The problem is that there is that
> + open window where the table could grow and we will end up
> + short on space. */
This needs to be fixed.
Currently, it seems that you just bail if this happens.
> +
> + buf_len = NETLBL_LEN_U32;
> + rcu_read_lock();
> + list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
> + if (iter->valid) {
> + doi_cnt += 1;
> + buf_len += 2 * NETLBL_LEN_U32;
> + }
> + rcu_read_unlock();
> +
> + skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_KERNEL);
> + if (skb == NULL)
> + return NULL;
> + buf = skb_put(skb, buf_len);
> + netlbl_putinc_u32(&buf, doi_cnt, &buf_len);
> +
> + rcu_read_lock();
> + list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
> + if (iter->valid) {
> + if (buf_len < 2 * NETLBL_LEN_U32) {
> + rcu_read_unlock();
> + kfree(skb);
> + return NULL;
> + }
> + netlbl_putinc_u32(&buf, iter->doi, &buf_len);
> + netlbl_putinc_u32(&buf, iter->type, &buf_len);
> + }
> + rcu_read_unlock();
> +
> + return skb;
> +}
> +
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 13:08 ` James Morris
@ 2006-07-15 13:11 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 13:11 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Sat, 15 Jul 2006, James Morris wrote:
> > + /* XXX - In both cases, this is kinda ugly as we have to go through
> > + the list once to determine how large of a buffer we need,
> > + drop the locks, allocate the buffer, grab the locks, and
> > + finally fill the buffer. The problem is that there is that
> > + open window where the table could grow and we will end up
> > + short on space. */
>
> This needs to be fixed.
Actually, you have many of these.
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 13:11 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 13:11 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Sat, 15 Jul 2006, James Morris wrote:
> > + /* XXX - In both cases, this is kinda ugly as we have to go through
> > + the list once to determine how large of a buffer we need,
> > + drop the locks, allocate the buffer, grab the locks, and
> > + finally fill the buffer. The problem is that there is that
> > + open window where the table could grow and we will end up
> > + short on space. */
>
> This needs to be fixed.
Actually, you have many of these.
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-15 13:15 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 13:15 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
> +{
> + if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
> + return -EINVAL;
> +
> + doi_def->valid = 1;
> + INIT_RCU_HEAD(&doi_def->rcu);
> + INIT_LIST_HEAD(&doi_def->dom_list);
> +
> + rcu_read_lock();
> + if (cipso_v4_doi_search(doi_def->doi) != NULL) {
> + rcu_read_unlock();
> + return -EEXIST;
> + }
> + spin_lock(&cipso_v4_doi_list_lock);
> + list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
> + spin_unlock(&cipso_v4_doi_list_lock);
> + rcu_read_unlock();
> +
> + return 0;
> +}
This is racy in that you can end up with duplicate entries in the list
between rcu_read_lock() and spin_lock().
You need to perform the search a second time under the spin_lock() as a
slow path to verify that an entry wasn't added, before adding a new one.
- James
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 13:15 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 13:15 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
> +{
> + if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
> + return -EINVAL;
> +
> + doi_def->valid = 1;
> + INIT_RCU_HEAD(&doi_def->rcu);
> + INIT_LIST_HEAD(&doi_def->dom_list);
> +
> + rcu_read_lock();
> + if (cipso_v4_doi_search(doi_def->doi) != NULL) {
> + rcu_read_unlock();
> + return -EEXIST;
> + }
> + spin_lock(&cipso_v4_doi_list_lock);
> + list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
> + spin_unlock(&cipso_v4_doi_list_lock);
> + rcu_read_unlock();
> +
> + return 0;
> +}
This is racy in that you can end up with duplicate entries in the list
between rcu_read_lock() and spin_lock().
You need to perform the search a second time under the spin_lock() as a
slow path to verify that an entry wasn't added, before adding a new one.
- James
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 2/7] NetLabel: core network changes
2006-07-14 23:34 ` James Morris
@ 2006-07-15 14:48 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 14:48 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Friday 14 July 2006 7:34 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +static inline void netlbl_put_hdr(unsigned char *buffer,
> > + const u32 msg_type,
> > + const u16 msg_len,
> > + const u16 msg_flags,
> > + const u32 msg_pid,
> > + const u32 msg_seq)
>
> Why are these parameters marked const? Seems to be common throughout the
> code.
>
I saw similar things in the kernel, but it looks like I was looking at
older, "less desirable" sections of code. Since it looks like you are still
churning through the patchset I'll hold off on submitting a new version based
on your comments until Monday.
Once again, thanks for your comments and patience.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 2/7] NetLabel: core network changes
@ 2006-07-15 14:48 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 14:48 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Friday 14 July 2006 7:34 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +static inline void netlbl_put_hdr(unsigned char *buffer,
> > + const u32 msg_type,
> > + const u16 msg_len,
> > + const u16 msg_flags,
> > + const u32 msg_pid,
> > + const u32 msg_seq)
>
> Why are these parameters marked const? Seems to be common throughout the
> code.
>
I saw similar things in the kernel, but it looks like I was looking at
older, "less desirable" sections of code. Since it looks like you are still
churning through the patchset I'll hold off on submitting a new version based
on your comments until Monday.
Once again, thanks for your comments and patience.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 2:03 ` James Morris
@ 2006-07-15 14:59 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 14:59 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Friday 14 July 2006 10:03 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +/**
> > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> >
> > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
>
> Can you use lib/bitmap.c instead?
I looked at using the routines in asm/bitopts.h at one point but decided not
too. Unfortunately, this was quite some time ago and I can't remember why I
chose not to use them right now. I'll look again and if it makes sense I'll
make the change.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 14:59 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 14:59 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Friday 14 July 2006 10:03 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +/**
> > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> >
> > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
>
> Can you use lib/bitmap.c instead?
I looked at using the routines in asm/bitopts.h at one point but decided not
too. Unfortunately, this was quite some time ago and I can't remember why I
chose not to use them right now. I'll look again and if it makes sense I'll
make the change.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 2:18 ` James Morris
@ 2006-07-15 15:03 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:03 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Friday 14 July 2006 10:18 pm, James Morris wrote:
>
> If you're in a permanent cleanup phase, why bother clearing all of these
> global variables and using temporary pointers?
>
> Why not just simply free each entry? The way this is coded makes it seem
> like you're unsure about the safety of the code.
>
> e.g. what could it possibly matter at this stage whether cipso_v4_cache is
> NULL or not?
>
> I think I asked this some time ago: what are the lifetime rules for this
> code as a loadable module? What if you arbitrarily rmmod it?
>
The answer to all of your comments in this email is that these appear to be
fragments from earlier thoughts about trying to make this code a loadable
module. Upon further reflection I realized that it probably doesn't make
much sense to have this code as a module for much of the same reason that
SELinux isn't setup as a loadable module.
As you have progressed through the patchset I'm sure you've realized by now
this code is not intended to function as a module. I'll clean these
leftovers up for the next version of the patchset.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 15:03 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:03 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Friday 14 July 2006 10:18 pm, James Morris wrote:
>
> If you're in a permanent cleanup phase, why bother clearing all of these
> global variables and using temporary pointers?
>
> Why not just simply free each entry? The way this is coded makes it seem
> like you're unsure about the safety of the code.
>
> e.g. what could it possibly matter at this stage whether cipso_v4_cache is
> NULL or not?
>
> I think I asked this some time ago: what are the lifetime rules for this
> code as a loadable module? What if you arbitrarily rmmod it?
>
The answer to all of your comments in this email is that these appear to be
fragments from earlier thoughts about trying to make this code a loadable
module. Upon further reflection I realized that it probably doesn't make
much sense to have this code as a module for much of the same reason that
SELinux isn't setup as a loadable module.
As you have progressed through the patchset I'm sure you've realized by now
this code is not intended to function as a module. I'll clean these
leftovers up for the next version of the patchset.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 13:08 ` James Morris
@ 2006-07-15 15:09 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:09 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 9:08 am, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
> > +{
> > + struct sk_buff *skb;
> > + unsigned char *buf;
> > + struct cipso_v4_doi *iter;
> > + u32 doi_cnt = 0;
> > + ssize_t buf_len;
> > +
> > + /* XXX - In both cases, this is kinda ugly as we have to go through
> > + the list once to determine how large of a buffer we need,
> > + drop the locks, allocate the buffer, grab the locks, and
> > + finally fill the buffer. The problem is that there is that
> > + open window where the table could grow and we will end up
> > + short on space. */
>
> This needs to be fixed.
The reasoning behind this was because I wanted to try and avoid holding the
rcu_read_lock() while allocating memory. I'll just change the allocation to
GFP_ATOMIC and keep the rcu lock.
> Currently, it seems that you just bail if this happens.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 15:09 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:09 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 9:08 am, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +struct sk_buff *cipso_v4_doi_dump_all(const size_t headroom)
> > +{
> > + struct sk_buff *skb;
> > + unsigned char *buf;
> > + struct cipso_v4_doi *iter;
> > + u32 doi_cnt = 0;
> > + ssize_t buf_len;
> > +
> > + /* XXX - In both cases, this is kinda ugly as we have to go through
> > + the list once to determine how large of a buffer we need,
> > + drop the locks, allocate the buffer, grab the locks, and
> > + finally fill the buffer. The problem is that there is that
> > + open window where the table could grow and we will end up
> > + short on space. */
>
> This needs to be fixed.
The reasoning behind this was because I wanted to try and avoid holding the
rcu_read_lock() while allocating memory. I'll just change the allocation to
GFP_ATOMIC and keep the rcu lock.
> Currently, it seems that you just bail if this happens.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 13:11 ` James Morris
@ 2006-07-15 15:13 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:13 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 9:11 am, James Morris wrote:
> On Sat, 15 Jul 2006, James Morris wrote:
> > > + /* XXX - In both cases, this is kinda ugly as we have to go through
> > > + the list once to determine how large of a buffer we need,
> > > + drop the locks, allocate the buffer, grab the locks, and
> > > + finally fill the buffer. The problem is that there is that
> > > + open window where the table could grow and we will end up
> > > + short on space. */
> >
> > This needs to be fixed.
>
> Actually, you have many of these.
I believe there are a total of four, the two you just found in
net/ipv4/cipso_ipv4.c and two more in net/netlabel/netlabel_domainhash.c. I
will fix all of them for the next submission of the patchset.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 15:13 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:13 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 9:11 am, James Morris wrote:
> On Sat, 15 Jul 2006, James Morris wrote:
> > > + /* XXX - In both cases, this is kinda ugly as we have to go through
> > > + the list once to determine how large of a buffer we need,
> > > + drop the locks, allocate the buffer, grab the locks, and
> > > + finally fill the buffer. The problem is that there is that
> > > + open window where the table could grow and we will end up
> > > + short on space. */
> >
> > This needs to be fixed.
>
> Actually, you have many of these.
I believe there are a total of four, the two you just found in
net/ipv4/cipso_ipv4.c and two more in net/netlabel/netlabel_domainhash.c. I
will fix all of them for the next submission of the patchset.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 13:15 ` James Morris
@ 2006-07-15 15:14 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:14 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 9:15 am, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
> > +{
> > + if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
> > + return -EINVAL;
> > +
> > + doi_def->valid = 1;
> > + INIT_RCU_HEAD(&doi_def->rcu);
> > + INIT_LIST_HEAD(&doi_def->dom_list);
> > +
> > + rcu_read_lock();
> > + if (cipso_v4_doi_search(doi_def->doi) != NULL) {
> > + rcu_read_unlock();
> > + return -EEXIST;
> > + }
> > + spin_lock(&cipso_v4_doi_list_lock);
> > + list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
> > + spin_unlock(&cipso_v4_doi_list_lock);
> > + rcu_read_unlock();
> > +
> > + return 0;
> > +}
>
> This is racy in that you can end up with duplicate entries in the list
> between rcu_read_lock() and spin_lock().
>
> You need to perform the search a second time under the spin_lock() as a
> slow path to verify that an entry wasn't added, before adding a new one.
>
Thanks, there are a few more place which might have similar problems - I'll
check those as well.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 15:14 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 15:14 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 9:15 am, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
> > +{
> > + if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
> > + return -EINVAL;
> > +
> > + doi_def->valid = 1;
> > + INIT_RCU_HEAD(&doi_def->rcu);
> > + INIT_LIST_HEAD(&doi_def->dom_list);
> > +
> > + rcu_read_lock();
> > + if (cipso_v4_doi_search(doi_def->doi) != NULL) {
> > + rcu_read_unlock();
> > + return -EEXIST;
> > + }
> > + spin_lock(&cipso_v4_doi_list_lock);
> > + list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
> > + spin_unlock(&cipso_v4_doi_list_lock);
> > + rcu_read_unlock();
> > +
> > + return 0;
> > +}
>
> This is racy in that you can end up with duplicate entries in the list
> between rcu_read_lock() and spin_lock().
>
> You need to perform the search a second time under the spin_lock() as a
> slow path to verify that an entry wasn't added, before adding a new one.
>
Thanks, there are a few more place which might have similar problems - I'll
check those as well.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-14 18:57 ` paul.moore
@ 2006-07-15 22:39 ` James Morris
-1 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 22:39 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +int cipso_v4_cache_add(const struct sk_buff *skb,
> + const struct netlbl_lsm_secattr *secattr)
> +{
It seems that this cache grows without bounds, correct?
Also, how do you handle the case of a change to a cached mapping?
- James
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 22:39 ` James Morris
0 siblings, 0 replies; 55+ messages in thread
From: James Morris @ 2006-07-15 22:39 UTC (permalink / raw)
To: Paul Moore; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> +int cipso_v4_cache_add(const struct sk_buff *skb,
> + const struct netlbl_lsm_secattr *secattr)
> +{
It seems that this cache grows without bounds, correct?
Also, how do you handle the case of a change to a cached mapping?
- James
--
James Morris
<jmorris@namei.org>
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 22:39 ` James Morris
@ 2006-07-15 23:26 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 23:26 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 6:39 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +int cipso_v4_cache_add(const struct sk_buff *skb,
> > + const struct netlbl_lsm_secattr *secattr)
> > +{
>
> It seems that this cache grows without bounds, correct?
Unless I messed something up the cache is limited to CIPSO_V4_CACHE_BUCKETS *
cipso_v4_cache_bucketsize; see the bottom half of cipso_v4_cache_add() for
details.
> Also, how do you handle the case of a change to a cached mapping?
Through the use of cipso_v4_cache_invalidate() which gets called on CIPSO DOI
deletes and SELinux policy [re]loads. It's a bit heavy handed in that it
clears the entire cache, but both events should be pretty rare in practice
and the extra work to clear specific cache entries probably isn't worth it.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-15 23:26 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-15 23:26 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, davem, sds, jmorris, pratt
On Saturday 15 July 2006 6:39 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +int cipso_v4_cache_add(const struct sk_buff *skb,
> > + const struct netlbl_lsm_secattr *secattr)
> > +{
>
> It seems that this cache grows without bounds, correct?
Unless I messed something up the cache is limited to CIPSO_V4_CACHE_BUCKETS *
cipso_v4_cache_bucketsize; see the bottom half of cipso_v4_cache_add() for
details.
> Also, how do you handle the case of a change to a cached mapping?
Through the use of cipso_v4_cache_invalidate() which gets called on CIPSO DOI
deletes and SELinux policy [re]loads. It's a bit heavy handed in that it
clears the entire cache, but both events should be pretty rare in practice
and the extra work to clear specific cache entries probably isn't worth it.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-15 2:03 ` James Morris
@ 2006-07-16 16:10 ` Paul Moore
-1 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-16 16:10 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Friday 14 July 2006 10:03 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +/**
> > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> >
> > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
>
> Can you use lib/bitmap.c instead?
Looking again at include/asm/bitops.h I think I now remember why I decided not
to use them in the first place. At least on the x86 side it looks like the
underlying btsl instruction may access an entire word when the bitmap may
only by a byte long. I imagine in the practical sense it probably wouldn't
be too serious of an issue, but there are too many corner cases here for me
to be able to say for certain so I'll defer to the experts here.
If it is safe it shouldn't be too difficult to use the set/clear_bit()
functions in cipso_v4_bitmap_setbit() and the find_*_bit() functions in
cipso_v4_bitmap_walk().
As an aside, I've made the other changes you suggested but I won't have a
chance to test them until Monday morning. Assuming all goes well during some
quick testing I'll post a new patchset around mid-day on Monday.
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-16 16:10 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-16 16:10 UTC (permalink / raw)
To: James Morris; +Cc: netdev, selinux, David S. Miller, sds, pratt
On Friday 14 July 2006 10:03 pm, James Morris wrote:
> On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > +/**
> > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> >
> > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
>
> Can you use lib/bitmap.c instead?
Looking again at include/asm/bitops.h I think I now remember why I decided not
to use them in the first place. At least on the x86 side it looks like the
underlying btsl instruction may access an entire word when the bitmap may
only by a byte long. I imagine in the practical sense it probably wouldn't
be too serious of an issue, but there are too many corner cases here for me
to be able to say for certain so I'll defer to the experts here.
If it is safe it shouldn't be too difficult to use the set/clear_bit()
functions in cipso_v4_bitmap_setbit() and the find_*_bit() functions in
cipso_v4_bitmap_walk().
As an aside, I've made the other changes you suggested but I won't have a
chance to test them until Monday morning. Assuming all goes well during some
quick testing I'll post a new patchset around mid-day on Monday.
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-16 16:10 ` Paul Moore
(?)
@ 2006-07-17 1:12 ` David Miller
2006-07-17 2:42 ` Paul Moore
-1 siblings, 1 reply; 55+ messages in thread
From: David Miller @ 2006-07-17 1:12 UTC (permalink / raw)
To: paul.moore; +Cc: jmorris, netdev, selinux, sds, pratt
From: Paul Moore <paul.moore@hp.com>
Date: Sun, 16 Jul 2006 12:10:44 -0400
> On Friday 14 July 2006 10:03 pm, James Morris wrote:
> > On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > > +/**
> > > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> > >
> > > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
> >
> > Can you use lib/bitmap.c instead?
>
> Looking again at include/asm/bitops.h I think I now remember why I
> decided not to use them in the first place.
lib/bitmap.c and the asm/bitops.h operations are two entirely
different animals.
Wrt. your asm/bitops.h concerns, is there any reason you cannot pad
out your bitmaps to be a modulo of "long" which is required for
those routines?
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-17 1:12 ` David Miller
@ 2006-07-17 2:42 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-17 2:42 UTC (permalink / raw)
To: David Miller; +Cc: jmorris, netdev, selinux, sds, pratt
On Sunday 16 July 2006 9:12 pm, David Miller wrote:
> From: Paul Moore <paul.moore@hp.com>
> Date: Sun, 16 Jul 2006 12:10:44 -0400
>
> > On Friday 14 July 2006 10:03 pm, James Morris wrote:
> > > On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > > > +/**
> > > > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> > > >
> > > > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
> > >
> > > Can you use lib/bitmap.c instead?
> >
> > Looking again at include/asm/bitops.h I think I now remember why I
> > decided not to use them in the first place.
>
> lib/bitmap.c and the asm/bitops.h operations are two entirely
> different animals.
I probably should have been more clear - I didn't see anything in lib/bitmap.c
(or include/linux/bitmap.h) that I think would have been useful. However,
include/linux/bitmap.h makes reference to asm/bitops.h which does have some
function which may have been useful if they didn't have the length
restrictions.
> Wrt. your asm/bitops.h concerns, is there any reason you cannot pad
> out your bitmaps to be a modulo of "long" which is required for
> those routines?
Right now I use both the bitmap_walk() and bitmap_setbit() routines to deal
with both CIPSO tags straight from the sk_buff as well as the internal bitmap
representation. Padding out the internal bitmaps would require some code
changes but there isn't much I can do about the packet I don't believe. True
it would probably be okay for most packets to assume you could access an
entire "long"s worth of memory but then again it would only take one evil
packet to start causing problems ...
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
@ 2006-07-17 2:42 ` Paul Moore
0 siblings, 0 replies; 55+ messages in thread
From: Paul Moore @ 2006-07-17 2:42 UTC (permalink / raw)
To: David Miller; +Cc: jmorris, netdev, selinux, sds, pratt
On Sunday 16 July 2006 9:12 pm, David Miller wrote:
> From: Paul Moore <paul.moore@hp.com>
> Date: Sun, 16 Jul 2006 12:10:44 -0400
>
> > On Friday 14 July 2006 10:03 pm, James Morris wrote:
> > > On Fri, 14 Jul 2006, paul.moore@hp.com wrote:
> > > > +/**
> > > > + * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
> > > >
> > > > + * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
> > >
> > > Can you use lib/bitmap.c instead?
> >
> > Looking again at include/asm/bitops.h I think I now remember why I
> > decided not to use them in the first place.
>
> lib/bitmap.c and the asm/bitops.h operations are two entirely
> different animals.
I probably should have been more clear - I didn't see anything in lib/bitmap.c
(or include/linux/bitmap.h) that I think would have been useful. However,
include/linux/bitmap.h makes reference to asm/bitops.h which does have some
function which may have been useful if they didn't have the length
restrictions.
> Wrt. your asm/bitops.h concerns, is there any reason you cannot pad
> out your bitmaps to be a modulo of "long" which is required for
> those routines?
Right now I use both the bitmap_walk() and bitmap_setbit() routines to deal
with both CIPSO tags straight from the sk_buff as well as the internal bitmap
representation. Padding out the internal bitmaps would require some code
changes but there isn't much I can do about the packet I don't believe. True
it would probably be okay for most packets to assume you could access an
entire "long"s worth of memory but then again it would only take one evil
packet to start causing problems ...
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
* Re: [PATCH 3/7] NetLabel: CIPSOv4 engine
2006-07-17 2:42 ` Paul Moore
(?)
@ 2006-07-17 2:53 ` David Miller
-1 siblings, 0 replies; 55+ messages in thread
From: David Miller @ 2006-07-17 2:53 UTC (permalink / raw)
To: paul.moore; +Cc: jmorris, netdev, selinux, sds, pratt
From: Paul Moore <paul.moore@hp.com>
Date: Sun, 16 Jul 2006 22:42:07 -0400
> Right now I use both the bitmap_walk() and bitmap_setbit() routines
> to deal with both CIPSO tags straight from the sk_buff as well as
> the internal bitmap representation. Padding out the internal
> bitmaps would require some code changes but there isn't much I can
> do about the packet I don't believe. True it would probably be okay
> for most packets to assume you could access an entire "long"s worth
> of memory but then again it would only take one evil packet to start
> causing problems ...
The access also has to be long aligned to be valid for
the asm/bitops.h interfaces as well, which is also
problematic for your case I guess.
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 5/7] NetLabel: CIPSOv4 and Unlabeled packet integration
2006-07-17 15:52 [PATCH 0/7] Updated patchset w/James' comments paul.moore
@ 2006-07-17 15:52 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-17 15:52 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
[-- Attachment #1: netlabel-core_components --]
[-- Type: text/plain, Size: 22369 bytes --]
Add CIPSO/IPv4 and unlabeled packet management to the NetLabel subsystem. The
CIPSO/IPv4 changes allow the configuration of CIPSO/IPv4 within the overall
NetLabel framework. The unlabeled packet changes allows NetLabel to pass
unlabeled packets without error.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/netlabel/netlabel_cipso_v4.c | 554 ++++++++++++++++++++++++++++++++++++++
net/netlabel/netlabel_unlabeled.c | 253 +++++++++++++++++
2 files changed, 807 insertions(+)
Index: linux-2.6.18-rc2/net/netlabel/netlabel_cipso_v4.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc2/net/netlabel/netlabel_cipso_v4.c
@@ -0,0 +1,554 @@
+/*
+ * NetLabel CIPSO/IPv4 Support
+ *
+ * This file defines the CIPSO/IPv4 functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+
+#include "netlabel_user.h"
+#include "netlabel_cipso_v4.h"
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_cipsov4_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_CIPSOV4_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * netlbl_cipsov4_doi_free - Frees a CIPSO V4 DOI definition
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to the DOI definition can be released
+ * safely.
+ *
+ */
+static void netlbl_cipsov4_doi_free(struct rcu_head *entry)
+{
+ struct cipso_v4_doi *ptr;
+
+ ptr = container_of(entry, struct cipso_v4_doi, rcu);
+ switch (ptr->type) {
+ case CIPSO_V4_MAP_STD:
+ kfree(ptr->map.std->lvl.cipso);
+ kfree(ptr->map.std->lvl.local);
+ kfree(ptr->map.std->cat.cipso);
+ kfree(ptr->map.std->cat.local);
+ break;
+ }
+ kfree(ptr);
+}
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_cipsov4_add_std - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_STD DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_std(u32 doi,
+ const unsigned char *msg,
+ size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ u32 num_lvls;
+ u32 num_cats;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+ u32 tmp_val_a;
+ u32 tmp_val_b;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_std_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_std_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std = kzalloc(sizeof(*doi_def->map.std), GFP_KERNEL);
+ if (doi_def->map.std == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_STD;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_std_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_std_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ if (msg_len < 6 * NETLBL_LEN_U32)
+ goto add_std_failure;
+
+ num_lvls = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_lvls == 0)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.local_size > CIPSO_V4_MAX_LOC_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local = kcalloc(doi_def->map.std->lvl.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->lvl.cipso_size = netlbl_getinc_u8(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.cipso_size > CIPSO_V4_MAX_REM_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.cipso = kcalloc(doi_def->map.std->lvl.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ num_cats = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ doi_def->map.std->cat.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.local_size > CIPSO_V4_MAX_LOC_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.local = kcalloc(doi_def->map.std->cat.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->cat.cipso_size = netlbl_getinc_u16(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.cipso_size > CIPSO_V4_MAX_REM_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.cipso = kcalloc(doi_def->map.std->cat.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ if (msg_len <
+ num_lvls * (NETLBL_LEN_U32 + NETLBL_LEN_U8) +
+ num_cats * (NETLBL_LEN_U32 + NETLBL_LEN_U16))
+ goto add_std_failure;
+
+ for (iter = 0; iter < doi_def->map.std->lvl.cipso_size; iter++)
+ doi_def->map.std->lvl.cipso[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->lvl.local_size; iter++)
+ doi_def->map.std->lvl.local[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->cat.cipso_size; iter++)
+ doi_def->map.std->cat.cipso[iter] = CIPSO_V4_INV_CAT;
+ for (iter = 0; iter < doi_def->map.std->cat.local_size; iter++)
+ doi_def->map.std->cat.local[iter] = CIPSO_V4_INV_CAT;
+
+ for (iter = 0; iter < num_lvls; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u8(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->lvl.local_size ||
+ tmp_val_b >= doi_def->map.std->lvl.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->lvl.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->lvl.local[tmp_val_a] = tmp_val_b;
+ }
+
+ for (iter = 0; iter < num_cats; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u16(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->cat.local_size ||
+ tmp_val_b >= doi_def->map.std->cat.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->cat.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->cat.local[tmp_val_a] = tmp_val_b;
+ }
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_std_failure;
+ return 0;
+
+add_std_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add_pass - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_PASS DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_pass(u32 doi,
+ const unsigned char *msg,
+ size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_pass_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_pass_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_pass_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_PASS;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_pass_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_pass_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_pass_failure;
+ return 0;
+
+add_pass_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add - Handle an ADD message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Create a new DOI definition based on the given ADD message and add it to the
+ * CIPSO V4 engine. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_add(struct sk_buff *skb, struct genl_info *info)
+
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ u32 map_type;
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto add_return;
+
+ if (msg_len < 2 * NETLBL_LEN_U32)
+ goto add_return;
+
+ doi = netlbl_getinc_u32(&msg, &msg_len);
+ map_type = netlbl_getinc_u32(&msg, &msg_len);
+ switch (map_type) {
+ case CIPSO_V4_MAP_STD:
+ ret_val = netlbl_cipsov4_add_std(doi, msg, msg_len);
+ break;
+ case CIPSO_V4_MAP_PASS:
+ ret_val = netlbl_cipsov4_add_pass(doi, msg, msg_len);
+ break;
+ }
+
+add_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ struct sk_buff *ans_skb;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32)
+ goto list_failure;
+
+ doi = netlbl_get_u32(msg);
+ ans_skb = cipso_v4_doi_dump(doi, NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto list_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LIST,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_listall - Handle a LISTALL message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LISTALL message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_listall(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ struct sk_buff *ans_skb;
+
+ ans_skb = cipso_v4_doi_dump_all(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto listall_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LISTALL,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto listall_failure;
+
+ return 0;
+
+listall_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_remove - Handle a REMOVE message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVE message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_remove(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto remove_return;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32) {
+ ret_val = -EINVAL;
+ goto remove_return;
+ }
+
+ doi = netlbl_get_u32(msg);
+ ret_val = cipso_v4_doi_remove(doi, netlbl_cipsov4_doi_free);
+
+remove_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_cipsov4_genl_c_add = {
+ .cmd = NLBL_CIPSOV4_C_ADD,
+ .flags = 0,
+ .doit = netlbl_cipsov4_add,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_remove = {
+ .cmd = NLBL_CIPSOV4_C_REMOVE,
+ .flags = 0,
+ .doit = netlbl_cipsov4_remove,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_list = {
+ .cmd = NLBL_CIPSOV4_C_LIST,
+ .flags = 0,
+ .doit = netlbl_cipsov4_list,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_listall = {
+ .cmd = NLBL_CIPSOV4_C_LISTALL,
+ .flags = 0,
+ .doit = netlbl_cipsov4_listall,
+ .dumpit = NULL,
+};
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_cipsov4_genl_init - Register the CIPSOv4 NetLabel component
+ *
+ * Description:
+ * Register the CIPSOv4 packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_cipsov4_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_cipsov4_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_add);
+ if (ret_val != 0)
+ return ret_val;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_remove);
+ if (ret_val != 0)
+ return ret_val;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_list);
+ if (ret_val != 0)
+ return ret_val;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_listall);
+ if (ret_val != 0)
+ return ret_val;
+
+ return 0;
+}
Index: linux-2.6.18-rc2/net/netlabel/netlabel_unlabeled.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc2/net/netlabel/netlabel_unlabeled.c
@@ -0,0 +1,253 @@
+/*
+ * NetLabel Unlabeled Support
+ *
+ * This file defines functions for dealing with unlabeled packets for the
+ * NetLabel system. The NetLabel system manages static and dynamic label
+ * mappings for network protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+#include <net/netlabel.h>
+#include <asm/bug.h>
+
+#include "netlabel_user.h"
+#include "netlabel_domainhash.h"
+#include "netlabel_unlabeled.h"
+
+/* Accept unlabeled packets flag */
+static atomic_t netlabel_unlabel_accept_flg = ATOMIC_INIT(0);
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_unlabel_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_UNLABELED_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_unlabel_accept - Handle an ACCEPT message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ACCEPT message and set the accept flag accordingly.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_accept(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ u32 value;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ return ret_val;
+
+ if (netlbl_netlink_payload_len(skb) == NETLBL_LEN_U32) {
+ value = netlbl_get_u32(msg);
+ if (value == 1 || value == 0) {
+ atomic_set(&netlabel_unlabel_accept_flg, value);
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+ }
+ }
+
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ EINVAL);
+ return -EINVAL;
+}
+
+/**
+ * netlbl_unlabel_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond with the current status.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ size_t data_size;
+ struct sk_buff *ans_skb;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto list_failure;
+
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_LIST,
+ data_size);
+ if (data == NULL)
+ goto list_failure;
+
+ netlbl_put_u32(data, atomic_read(&netlabel_unlabel_accept_flg));
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_unlabel_genl_c_accept = {
+ .cmd = NLBL_UNLABEL_C_ACCEPT,
+ .flags = 0,
+ .doit = netlbl_unlabel_accept,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_unlabel_genl_c_list = {
+ .cmd = NLBL_UNLABEL_C_LIST,
+ .flags = 0,
+ .doit = netlbl_unlabel_list,
+ .dumpit = NULL,
+};
+
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_unlabel_genl_init - Register the Unlabeled NetLabel component
+ *
+ * Description:
+ * Register the unlabeled packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_unlabel_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_unlabel_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_accept);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_list);
+ if (ret_val != 0)
+ return ret_val;
+
+ return 0;
+}
+
+/*
+ * NetLabel KAPI Hooks
+ */
+
+/**
+ * netlbl_unlabel_getattr - Get the security attributes for an unlabled packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Determine the security attributes, if any, for an unlabled packet and return
+ * them in @secattr. Returns zero on success and negative values on failure.
+ *
+ */
+int netlbl_unlabel_getattr(struct netlbl_lsm_secattr *secattr)
+{
+ if (atomic_read(&netlabel_unlabel_accept_flg) == 1) {
+ memset(secattr, 0, sizeof(*secattr));
+ return 0;
+ }
+
+ return -ENOMSG;
+}
+
+/**
+ * netlbl_unlabel_defconf - Set the default config to allow unlabeled packets
+ *
+ * Description:
+ * Set the default NetLabel configuration to allow incoming unlabeled packets
+ * and to send unlabeled network traffic by default.
+ *
+ */
+int netlbl_unlabel_defconf(void)
+{
+ int ret_val;
+ struct netlbl_dom_map *entry;
+
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL)
+ return -ENOMEM;
+ entry->type = NETLBL_NLTYPE_UNLABELED;
+ ret_val = netlbl_domhsh_add_default(entry);
+ if (ret_val != 0)
+ return ret_val;
+
+ atomic_set(&netlabel_unlabel_accept_flg, 1);
+
+ return 0;
+}
--
paul moore
linux security @ hp
^ permalink raw reply [flat|nested] 55+ messages in thread
* [PATCH 5/7] NetLabel: CIPSOv4 and Unlabeled packet integration
@ 2006-07-17 15:52 ` paul.moore
0 siblings, 0 replies; 55+ messages in thread
From: paul.moore @ 2006-07-17 15:52 UTC (permalink / raw)
To: netdev, selinux; +Cc: davem, sds, jmorris, pratt, Paul Moore
Add CIPSO/IPv4 and unlabeled packet management to the NetLabel subsystem. The
CIPSO/IPv4 changes allow the configuration of CIPSO/IPv4 within the overall
NetLabel framework. The unlabeled packet changes allows NetLabel to pass
unlabeled packets without error.
Signed-off-by: Paul Moore <paul.moore@hp.com>
---
net/netlabel/netlabel_cipso_v4.c | 554 ++++++++++++++++++++++++++++++++++++++
net/netlabel/netlabel_unlabeled.c | 253 +++++++++++++++++
2 files changed, 807 insertions(+)
Index: linux-2.6.18-rc2/net/netlabel/netlabel_cipso_v4.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc2/net/netlabel/netlabel_cipso_v4.c
@@ -0,0 +1,554 @@
+/*
+ * NetLabel CIPSO/IPv4 Support
+ *
+ * This file defines the CIPSO/IPv4 functions for the NetLabel system. The
+ * NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+
+#include "netlabel_user.h"
+#include "netlabel_cipso_v4.h"
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_cipsov4_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_CIPSOV4_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * netlbl_cipsov4_doi_free - Frees a CIPSO V4 DOI definition
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to the DOI definition can be released
+ * safely.
+ *
+ */
+static void netlbl_cipsov4_doi_free(struct rcu_head *entry)
+{
+ struct cipso_v4_doi *ptr;
+
+ ptr = container_of(entry, struct cipso_v4_doi, rcu);
+ switch (ptr->type) {
+ case CIPSO_V4_MAP_STD:
+ kfree(ptr->map.std->lvl.cipso);
+ kfree(ptr->map.std->lvl.local);
+ kfree(ptr->map.std->cat.cipso);
+ kfree(ptr->map.std->cat.local);
+ break;
+ }
+ kfree(ptr);
+}
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_cipsov4_add_std - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_STD DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_std(u32 doi,
+ const unsigned char *msg,
+ size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ u32 num_lvls;
+ u32 num_cats;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+ u32 tmp_val_a;
+ u32 tmp_val_b;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_std_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_std_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std = kzalloc(sizeof(*doi_def->map.std), GFP_KERNEL);
+ if (doi_def->map.std == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_STD;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_std_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_std_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ if (msg_len < 6 * NETLBL_LEN_U32)
+ goto add_std_failure;
+
+ num_lvls = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_lvls == 0)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.local_size > CIPSO_V4_MAX_LOC_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.local = kcalloc(doi_def->map.std->lvl.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->lvl.cipso_size = netlbl_getinc_u8(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->lvl.cipso_size > CIPSO_V4_MAX_REM_LVLS)
+ goto add_std_failure;
+ doi_def->map.std->lvl.cipso = kcalloc(doi_def->map.std->lvl.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->lvl.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ num_cats = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ doi_def->map.std->cat.local_size = netlbl_getinc_u32(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.local_size > CIPSO_V4_MAX_LOC_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.local = kcalloc(doi_def->map.std->cat.local_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.local == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+ doi_def->map.std->cat.cipso_size = netlbl_getinc_u16(&msg_ptr,
+ &msg_len);
+ if (doi_def->map.std->cat.cipso_size > CIPSO_V4_MAX_REM_CATS)
+ goto add_std_failure;
+ doi_def->map.std->cat.cipso = kcalloc(doi_def->map.std->cat.cipso_size,
+ sizeof(u32),
+ GFP_KERNEL);
+ if (doi_def->map.std->cat.cipso == NULL) {
+ ret_val = -ENOMEM;
+ goto add_std_failure;
+ }
+
+ if (msg_len <
+ num_lvls * (NETLBL_LEN_U32 + NETLBL_LEN_U8) +
+ num_cats * (NETLBL_LEN_U32 + NETLBL_LEN_U16))
+ goto add_std_failure;
+
+ for (iter = 0; iter < doi_def->map.std->lvl.cipso_size; iter++)
+ doi_def->map.std->lvl.cipso[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->lvl.local_size; iter++)
+ doi_def->map.std->lvl.local[iter] = CIPSO_V4_INV_LVL;
+ for (iter = 0; iter < doi_def->map.std->cat.cipso_size; iter++)
+ doi_def->map.std->cat.cipso[iter] = CIPSO_V4_INV_CAT;
+ for (iter = 0; iter < doi_def->map.std->cat.local_size; iter++)
+ doi_def->map.std->cat.local[iter] = CIPSO_V4_INV_CAT;
+
+ for (iter = 0; iter < num_lvls; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u8(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->lvl.local_size ||
+ tmp_val_b >= doi_def->map.std->lvl.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->lvl.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->lvl.local[tmp_val_a] = tmp_val_b;
+ }
+
+ for (iter = 0; iter < num_cats; iter++) {
+ tmp_val_a = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ tmp_val_b = netlbl_getinc_u16(&msg_ptr, &msg_len);
+
+ if (tmp_val_a >= doi_def->map.std->cat.local_size ||
+ tmp_val_b >= doi_def->map.std->cat.cipso_size)
+ goto add_std_failure;
+
+ doi_def->map.std->cat.cipso[tmp_val_b] = tmp_val_a;
+ doi_def->map.std->cat.local[tmp_val_a] = tmp_val_b;
+ }
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_std_failure;
+ return 0;
+
+add_std_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add_pass - Adds a CIPSO V4 DOI definition
+ * @doi: the DOI value
+ * @msg: the ADD message data
+ * @msg_size: the size of the ADD message buffer
+ *
+ * Description:
+ * Create a new CIPSO_V4_MAP_PASS DOI definition based on the given ADD message
+ * and add it to the CIPSO V4 engine. Return zero on success and non-zero on
+ * error.
+ *
+ */
+static int netlbl_cipsov4_add_pass(u32 doi,
+ const unsigned char *msg,
+ size_t msg_size)
+{
+ int ret_val = -EPERM;
+ unsigned char *msg_ptr = (unsigned char *)msg;
+ ssize_t msg_len = msg_size;
+ u32 num_tags;
+ struct cipso_v4_doi *doi_def = NULL;
+ u32 iter;
+
+ if (msg_len < NETLBL_LEN_U32)
+ goto add_pass_failure;
+ num_tags = netlbl_getinc_u32(&msg_ptr, &msg_len);
+ if (num_tags == 0 || num_tags > CIPSO_V4_TAG_MAXCNT)
+ goto add_pass_failure;
+
+ doi_def = kmalloc(sizeof(*doi_def), GFP_KERNEL);
+ if (doi_def == NULL) {
+ ret_val = -ENOMEM;
+ goto add_pass_failure;
+ }
+ doi_def->type = CIPSO_V4_MAP_PASS;
+
+ for (iter = 0; iter < num_tags; iter++) {
+ if (msg_len < NETLBL_LEN_U8)
+ goto add_pass_failure;
+ doi_def->tags[iter] = netlbl_getinc_u8(&msg_ptr, &msg_len);
+ switch (doi_def->tags[iter]) {
+ case CIPSO_V4_TAG_RBITMAP:
+ break;
+ default:
+ goto add_pass_failure;
+ }
+ }
+ if (iter < CIPSO_V4_TAG_MAXCNT)
+ doi_def->tags[iter] = CIPSO_V4_TAG_INVALID;
+
+ doi_def->doi = doi;
+ ret_val = cipso_v4_doi_add(doi_def);
+ if (ret_val != 0)
+ goto add_pass_failure;
+ return 0;
+
+add_pass_failure:
+ if (doi_def)
+ netlbl_cipsov4_doi_free(&doi_def->rcu);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_add - Handle an ADD message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Create a new DOI definition based on the given ADD message and add it to the
+ * CIPSO V4 engine. Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_add(struct sk_buff *skb, struct genl_info *info)
+
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ u32 map_type;
+ ssize_t msg_len = netlbl_netlink_payload_len(skb);
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto add_return;
+
+ if (msg_len < 2 * NETLBL_LEN_U32)
+ goto add_return;
+
+ doi = netlbl_getinc_u32(&msg, &msg_len);
+ map_type = netlbl_getinc_u32(&msg, &msg_len);
+ switch (map_type) {
+ case CIPSO_V4_MAP_STD:
+ ret_val = netlbl_cipsov4_add_std(doi, msg, msg_len);
+ break;
+ case CIPSO_V4_MAP_PASS:
+ ret_val = netlbl_cipsov4_add_pass(doi, msg, msg_len);
+ break;
+ }
+
+add_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ struct sk_buff *ans_skb;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32)
+ goto list_failure;
+
+ doi = netlbl_get_u32(msg);
+ ans_skb = cipso_v4_doi_dump(doi, NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto list_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LIST,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_listall - Handle a LISTALL message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LISTALL message and respond accordingly. Returns
+ * zero on success and negative values on error.
+ *
+ */
+static int netlbl_cipsov4_listall(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -EINVAL;
+ struct sk_buff *ans_skb;
+
+ ans_skb = cipso_v4_doi_dump_all(NLMSG_SPACE(GENL_HDRLEN));
+ if (ans_skb == NULL) {
+ ret_val = -ENOMEM;
+ goto listall_failure;
+ }
+ netlbl_netlink_hdr_push(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_LISTALL,
+ ans_skb->len);
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto listall_failure;
+
+ return 0;
+
+listall_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/**
+ * netlbl_cipsov4_remove - Handle a REMOVE message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated REMOVE message and respond accordingly. Returns
+ * zero on success, negative values on failure.
+ *
+ */
+static int netlbl_cipsov4_remove(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ u32 doi;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ goto remove_return;
+
+ if (netlbl_netlink_payload_len(skb) != NETLBL_LEN_U32) {
+ ret_val = -EINVAL;
+ goto remove_return;
+ }
+
+ doi = netlbl_get_u32(msg);
+ ret_val = cipso_v4_doi_remove(doi, netlbl_cipsov4_doi_free);
+
+remove_return:
+ netlbl_netlink_send_ack(info,
+ netlbl_cipsov4_gnl_family.id,
+ NLBL_CIPSOV4_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_cipsov4_genl_c_add = {
+ .cmd = NLBL_CIPSOV4_C_ADD,
+ .flags = 0,
+ .doit = netlbl_cipsov4_add,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_remove = {
+ .cmd = NLBL_CIPSOV4_C_REMOVE,
+ .flags = 0,
+ .doit = netlbl_cipsov4_remove,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_list = {
+ .cmd = NLBL_CIPSOV4_C_LIST,
+ .flags = 0,
+ .doit = netlbl_cipsov4_list,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_cipsov4_genl_c_listall = {
+ .cmd = NLBL_CIPSOV4_C_LISTALL,
+ .flags = 0,
+ .doit = netlbl_cipsov4_listall,
+ .dumpit = NULL,
+};
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_cipsov4_genl_init - Register the CIPSOv4 NetLabel component
+ *
+ * Description:
+ * Register the CIPSOv4 packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_cipsov4_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_cipsov4_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_add);
+ if (ret_val != 0)
+ return ret_val;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_remove);
+ if (ret_val != 0)
+ return ret_val;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_list);
+ if (ret_val != 0)
+ return ret_val;
+ ret_val = genl_register_ops(&netlbl_cipsov4_gnl_family,
+ &netlbl_cipsov4_genl_c_listall);
+ if (ret_val != 0)
+ return ret_val;
+
+ return 0;
+}
Index: linux-2.6.18-rc2/net/netlabel/netlabel_unlabeled.c
===================================================================
--- /dev/null
+++ linux-2.6.18-rc2/net/netlabel/netlabel_unlabeled.c
@@ -0,0 +1,253 @@
+/*
+ * NetLabel Unlabeled Support
+ *
+ * This file defines functions for dealing with unlabeled packets for the
+ * NetLabel system. The NetLabel system manages static and dynamic label
+ * mappings for network protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/socket.h>
+#include <linux/string.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+#include <net/netlabel.h>
+#include <asm/bug.h>
+
+#include "netlabel_user.h"
+#include "netlabel_domainhash.h"
+#include "netlabel_unlabeled.h"
+
+/* Accept unlabeled packets flag */
+static atomic_t netlabel_unlabel_accept_flg = ATOMIC_INIT(0);
+
+/* NetLabel Generic NETLINK CIPSOv4 family */
+static struct genl_family netlbl_unlabel_gnl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = NETLBL_NLTYPE_UNLABELED_NAME,
+ .version = NETLBL_PROTO_VERSION,
+ .maxattr = 0,
+};
+
+
+/*
+ * NetLabel Command Handlers
+ */
+
+/**
+ * netlbl_unlabel_accept - Handle an ACCEPT message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated ACCEPT message and set the accept flag accordingly.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_accept(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val;
+ unsigned char *msg = netlbl_netlink_payload_data(skb);
+ u32 value;
+
+ ret_val = netlbl_netlink_cap_check(skb, CAP_NET_ADMIN);
+ if (ret_val != 0)
+ return ret_val;
+
+ if (netlbl_netlink_payload_len(skb) == NETLBL_LEN_U32) {
+ value = netlbl_get_u32(msg);
+ if (value == 1 || value == 0) {
+ atomic_set(&netlabel_unlabel_accept_flg, value);
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ NETLBL_E_OK);
+ return 0;
+ }
+ }
+
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ EINVAL);
+ return -EINVAL;
+}
+
+/**
+ * netlbl_unlabel_list - Handle a LIST message
+ * @skb: the NETLINK buffer
+ * @info: the Generic NETLINK info block
+ *
+ * Description:
+ * Process a user generated LIST message and respond with the current status.
+ * Returns zero on success, negative values on failure.
+ *
+ */
+static int netlbl_unlabel_list(struct sk_buff *skb, struct genl_info *info)
+{
+ int ret_val = -ENOMEM;
+ size_t data_size;
+ struct sk_buff *ans_skb;
+ unsigned char *data;
+
+ data_size = GENL_HDRLEN + NETLBL_LEN_U32;
+ ans_skb = netlbl_netlink_alloc_skb(0, data_size, GFP_KERNEL);
+ if (ans_skb == NULL)
+ goto list_failure;
+
+ data = netlbl_netlink_hdr_put(ans_skb,
+ info->snd_pid,
+ 0,
+ 0,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_LIST,
+ data_size);
+ if (data == NULL)
+ goto list_failure;
+
+ netlbl_put_u32(data, atomic_read(&netlabel_unlabel_accept_flg));
+
+ ret_val = netlbl_netlink_snd(ans_skb, info->snd_pid);
+ if (ret_val != 0)
+ goto list_failure;
+
+ return 0;
+
+list_failure:
+ netlbl_netlink_send_ack(info,
+ netlbl_unlabel_gnl_family.id,
+ NLBL_UNLABEL_C_ACK,
+ -ret_val);
+ return ret_val;
+}
+
+
+/*
+ * NetLabel Generic NETLINK Command Definitions
+ */
+
+static struct genl_ops netlbl_unlabel_genl_c_accept = {
+ .cmd = NLBL_UNLABEL_C_ACCEPT,
+ .flags = 0,
+ .doit = netlbl_unlabel_accept,
+ .dumpit = NULL,
+};
+
+static struct genl_ops netlbl_unlabel_genl_c_list = {
+ .cmd = NLBL_UNLABEL_C_LIST,
+ .flags = 0,
+ .doit = netlbl_unlabel_list,
+ .dumpit = NULL,
+};
+
+
+/*
+ * NetLabel Generic NETLINK Protocol Functions
+ */
+
+/**
+ * netlbl_unlabel_genl_init - Register the Unlabeled NetLabel component
+ *
+ * Description:
+ * Register the unlabeled packet NetLabel component with the Generic NETLINK
+ * mechanism. Returns zero on success, negative values on failure.
+ *
+ */
+int netlbl_unlabel_genl_init(void)
+{
+ int ret_val;
+
+ ret_val = genl_register_family(&netlbl_unlabel_gnl_family);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_accept);
+ if (ret_val != 0)
+ return ret_val;
+
+ ret_val = genl_register_ops(&netlbl_unlabel_gnl_family,
+ &netlbl_unlabel_genl_c_list);
+ if (ret_val != 0)
+ return ret_val;
+
+ return 0;
+}
+
+/*
+ * NetLabel KAPI Hooks
+ */
+
+/**
+ * netlbl_unlabel_getattr - Get the security attributes for an unlabled packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Determine the security attributes, if any, for an unlabled packet and return
+ * them in @secattr. Returns zero on success and negative values on failure.
+ *
+ */
+int netlbl_unlabel_getattr(struct netlbl_lsm_secattr *secattr)
+{
+ if (atomic_read(&netlabel_unlabel_accept_flg) == 1) {
+ memset(secattr, 0, sizeof(*secattr));
+ return 0;
+ }
+
+ return -ENOMSG;
+}
+
+/**
+ * netlbl_unlabel_defconf - Set the default config to allow unlabeled packets
+ *
+ * Description:
+ * Set the default NetLabel configuration to allow incoming unlabeled packets
+ * and to send unlabeled network traffic by default.
+ *
+ */
+int netlbl_unlabel_defconf(void)
+{
+ int ret_val;
+ struct netlbl_dom_map *entry;
+
+ entry = kzalloc(sizeof(*entry), GFP_KERNEL);
+ if (entry == NULL)
+ return -ENOMEM;
+ entry->type = NETLBL_NLTYPE_UNLABELED;
+ ret_val = netlbl_domhsh_add_default(entry);
+ if (ret_val != 0)
+ return ret_val;
+
+ atomic_set(&netlabel_unlabel_accept_flg, 1);
+
+ return 0;
+}
--
paul moore
linux security @ hp
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.
^ permalink raw reply [flat|nested] 55+ messages in thread
end of thread, other threads:[~2006-07-17 15:58 UTC | newest]
Thread overview: 55+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2006-07-14 18:57 [PATCH 0/7] Latest NetLabel patch for 2.6.19 paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 18:57 ` [PATCH 1/7] NetLabel: documentation paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 18:57 ` [PATCH 2/7] NetLabel: core network changes paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 23:34 ` James Morris
2006-07-14 23:34 ` James Morris
2006-07-14 23:36 ` David Miller
2006-07-15 14:48 ` Paul Moore
2006-07-15 14:48 ` Paul Moore
2006-07-14 18:57 ` [PATCH 3/7] NetLabel: CIPSOv4 engine paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 23:56 ` James Morris
2006-07-14 23:56 ` James Morris
2006-07-15 2:03 ` James Morris
2006-07-15 2:03 ` James Morris
2006-07-15 14:59 ` Paul Moore
2006-07-15 14:59 ` Paul Moore
2006-07-16 16:10 ` Paul Moore
2006-07-16 16:10 ` Paul Moore
2006-07-17 1:12 ` David Miller
2006-07-17 2:42 ` Paul Moore
2006-07-17 2:42 ` Paul Moore
2006-07-17 2:53 ` David Miller
2006-07-15 2:18 ` James Morris
2006-07-15 2:18 ` James Morris
2006-07-15 15:03 ` Paul Moore
2006-07-15 15:03 ` Paul Moore
2006-07-15 13:08 ` James Morris
2006-07-15 13:08 ` James Morris
2006-07-15 13:11 ` James Morris
2006-07-15 13:11 ` James Morris
2006-07-15 15:13 ` Paul Moore
2006-07-15 15:13 ` Paul Moore
2006-07-15 15:09 ` Paul Moore
2006-07-15 15:09 ` Paul Moore
2006-07-15 13:15 ` James Morris
2006-07-15 13:15 ` James Morris
2006-07-15 15:14 ` Paul Moore
2006-07-15 15:14 ` Paul Moore
2006-07-15 22:39 ` James Morris
2006-07-15 22:39 ` James Morris
2006-07-15 23:26 ` Paul Moore
2006-07-15 23:26 ` Paul Moore
2006-07-14 18:57 ` [PATCH 4/7] NetLabel: core NetLabel subsystem paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 18:57 ` [PATCH 5/7] NetLabel: CIPSOv4 and Unlabeled packet integration paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 18:57 ` [PATCH 6/7] NetLabel: SELinux support paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-14 18:57 ` [PATCH 7/7] NetLabel: tie NetLabel into the Kconfig system paul.moore
2006-07-14 18:57 ` paul.moore
2006-07-17 15:52 [PATCH 0/7] Updated patchset w/James' comments paul.moore
2006-07-17 15:52 ` [PATCH 5/7] NetLabel: CIPSOv4 and Unlabeled packet integration paul.moore
2006-07-17 15:52 ` paul.moore
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.