Tobias Brunner [Fri, 25 Apr 2014 16:58:55 +0000 (18:58 +0200)]
child-cfg: Fix removal of redundant traffic selectors
We have to make sure we compare every selected traffic selector with every
other in the list.
Fixes #577.
Tobias Brunner [Fri, 25 Apr 2014 12:38:46 +0000 (14:38 +0200)]
android: New release based on 5.1.3
Also links OpenSSL statically and doesn't limit the number of packets
during EAP-TTLS.
Tobias Brunner [Fri, 25 Apr 2014 12:26:01 +0000 (14:26 +0200)]
libcharon: Added AEAD sources of libtls to Android.mk
Tobias Brunner [Fri, 25 Apr 2014 12:19:09 +0000 (14:19 +0200)]
libimcv: Updated Android.mk
Tobias Brunner [Thu, 27 Mar 2014 17:04:18 +0000 (18:04 +0100)]
android: Use static version of libcrypto
System.loadLibrary() searches in system directories first (at least in
recent releases), that is, our own build wouldn't actually get used.
Tobias Brunner [Fri, 25 Apr 2014 09:28:52 +0000 (11:28 +0200)]
tun-device: Use SIOCAIFADDR to set IP address on FreeBSD 10
FreeBSD 10 deprecated the SIOCSIFADDR etc. commands, so we use this
newer command to set the address and netmask. A destination address
is now also required.
Fixes #566.
Tobias Brunner [Thu, 24 Apr 2014 15:55:25 +0000 (17:55 +0200)]
Merge branch 'atomic-ref'
Adds support for GCC's __atomic* built-ins and improves the performance
of logging (for ignored log levels) and half-open IKE_SA checking under
high loads.
Also fixes two potential race conditions in the load-tester plugin.
Tobias Brunner [Thu, 17 Apr 2014 08:47:32 +0000 (10:47 +0200)]
bus: Add a fast-path if log messages don't have to be logged
For some rwlock_t implementations acquiring the read lock could be quite
expensive even if there are no writers (e.g. because the implementation
requires acquiring a mutex to check for writers) particularly if the
lock is highly contended, like it is for the vlog() method.
Christophe Gouault [Tue, 8 Apr 2014 15:11:14 +0000 (17:11 +0200)]
load-tester: Fix race condition issuing same SPI
Due to an unprotected incrementation, two load-tester initiators occasionally
use the same SPI under high load, and hence generate 2 IPsec SAs with the same
identifier. The responder IPsec stack will refuse to configure the second SA.
Use an atomic incrementation to avoid this race condition.
Signed-off-by: Christophe Gouault <christophe.gouault@6wind.com>
Christophe Gouault [Tue, 8 Apr 2014 15:11:13 +0000 (17:11 +0200)]
load-tester: Fix race condition issuing same identity
Due to an unprotected incrementation, two load-tester initiators occasionally
use the same identifier under high load. The responder typically drops one of
the connections.
Use an atomic incrementation to avoid this race condition.
Signed-off-by: Christophe Gouault <christophe.gouault@6wind.com>
Tobias Brunner [Fri, 11 Apr 2014 14:23:39 +0000 (16:23 +0200)]
ike-sa-manager: Improve scalability of half-open IKE_SA checking
This patch is based on one by Christoph Gouault.
Currently, to count the total number of half_open IKE_SAs,
get_half_open_count sums up the count of each segment in the SA hash
table (acquiring a lock for each segment). This procedure does not scale
well when the number of segments increases, as the method is called for
each new negotiation.
Instead, lets maintain a global atomic counter.
This optimization allows the use of big values for charon.ikesa_table_size
and charon.ikesa_table_segments.
Tobias Brunner [Fri, 11 Apr 2014 14:07:32 +0000 (16:07 +0200)]
utils: Use GCC's __atomic built-ins if available
These are available since GCC 4.7 and will eventually replace the __sync
operations. They support the memory model defined by C++11. For instance,
by using __ATOMIC_RELAXED for some operations on the reference counters we
can avoid memory barriers, which are required by __sync operations (whose
memory model essentially is __ATOMIC_SEQ_CST).
Tobias Brunner [Fri, 11 Apr 2014 13:13:22 +0000 (15:13 +0200)]
utils: Add ref_cur() to retrieve the current value of a reference counter
On many architectures it is safe to read the value directly (those
using cache coherency protocols, and with atomic loads for 32-bit
values) but it is not if that's not the case or if we ever decide to
make refcount_t 64-bit (load not atomic on x86).
So make sure the operation is actually atomic and that users do not
have to care about the size of refcount_t.
Tobias Brunner [Tue, 22 Apr 2014 14:27:41 +0000 (16:27 +0200)]
testing: Added pfkey/compress test case
Francois ten Krooden [Tue, 22 Apr 2014 13:34:41 +0000 (15:34 +0200)]
kernel-pfkey: Added IPComp support
- get_cpi function was implemented to retrieve a CPI from the kernel.
- add_sa/update_sa/del_sa were updated to accommodate for IPComp SA.
- Updated add_policy_internal to update the SPD to support IPComp.
Martin Willi [Thu, 24 Apr 2014 13:45:30 +0000 (15:45 +0200)]
packages: New Debian network-manager-strongswan release
Martin Willi [Thu, 24 Apr 2014 13:40:56 +0000 (15:40 +0200)]
packages: Hand over network-manager-strongswan debian package maintenance
Martin Willi [Thu, 24 Apr 2014 13:34:50 +0000 (15:34 +0200)]
packages: Use charon-nm in network-manager-strongswan debian package
Martin Willi [Thu, 24 Apr 2014 13:38:21 +0000 (15:38 +0200)]
nm: Bump NetworkManager plugin version to 1.3.1
Andreas Steffen [Thu, 24 Apr 2014 07:08:07 +0000 (09:08 +0200)]
pacman.sh now fetches Ubuntu 14.04 security updates
Martin Willi [Thu, 17 Apr 2014 08:14:49 +0000 (10:14 +0200)]
Merge branch 'reauth-collision'
Fixes two collisions between IKE_SA re-authentication and CHILD_SA rekeying.
Martin Willi [Thu, 10 Apr 2014 09:31:17 +0000 (11:31 +0200)]
ike: Delay actively initiated reauthentication when other exchanges in progress
If any other IKE or CHILD_SA operation takes places, we should not start
initiating reauthentication to avoid any potential races.
Martin Willi [Thu, 10 Apr 2014 09:25:32 +0000 (11:25 +0200)]
ikev2: Reject CHILD_SA creation/rekeying while deleting an IKE_SA
If one peer starts reauthentication by deleting the IKE_SA, while the other
starts CHILD_SA rekeying, we run in a race condition. To avoid it, temporarily
reject the rekey attempt while we are in the IKE_SA deleting state.
RFC 4306/5996 is not exactly clear about this collision, but it should be safe
to reject CHILD_SA rekeying during this stage, as the reauth will re-trigger the
CHILD_SA. For non-rekeying CHILD_SA creations, it's up to the peer to retry
establishing the CHILD_SA on the reauthenticated IKE_SA.
Martin Willi [Thu, 10 Apr 2014 08:24:34 +0000 (10:24 +0200)]
ikev2: Apply extensions and conditions before starting rekeying
The extensions and conditions apply to the rekeyed IKE_SA as well, so we should
migrate them. Especially when using algorithms from private space, we need
EXT_STRONGSWAN to properly select these algorithms during IKE rekeying.
Martin Willi [Thu, 10 Apr 2014 08:21:32 +0000 (10:21 +0200)]
ikev2: Add inherit_pre() to apply config and hosts before IKE_SA rekeying
Martin Willi [Mon, 14 Apr 2014 12:42:27 +0000 (14:42 +0200)]
ikev1: Add an option to accept unencrypted ID/HASH payloads
Even in Main Mode, some Sonicwall boxes seem to send ID/HASH payloads in
unencrypted form, probably to allow PSK lookup based on the ID payloads. We
by default reject that, but accept it if the
charon.accept_unencrypted_mainmode_messages option is set in strongswan.conf.
Initial patch courtesy of Paul Stewart.
Tobias Brunner [Tue, 15 Apr 2014 14:00:47 +0000 (16:00 +0200)]
ikev2: Fix reauthentication if peer assigns a different virtual IP
Before this change a reqid set on the create_child_t task was used as
indicator of the CHILD_SA being rekeyed. Only if that was not the case
would the local traffic selector be changed to 0.0.0.0/0|::/0 (as we
don't know which virtual IP the gateway will eventually assign).
On the other hand, in case of a rekeying the VIP is expected to remain
the same, so the local TS would simply equal the VIP.
Since
c949a4d5016e33c5 reauthenticated CHILD_SAs also have the reqid
set. Which meant that the local TS would contain the previously
assigned VIP, basically rendering the gateway unable to assign a
different VIP to the client as the resulting TS would not match
the client's proposal anymore.
Fixes #553.
Andreas Steffen [Tue, 15 Apr 2014 08:04:27 +0000 (10:04 +0200)]
Added NEWS for 5.2.0dr1
Andreas Steffen [Tue, 15 Apr 2014 06:55:11 +0000 (08:55 +0200)]
Handle tag separators
Andreas Steffen [Mon, 14 Apr 2014 20:52:26 +0000 (22:52 +0200)]
Renewed expired user certificate
Andreas Steffen [Mon, 14 Apr 2014 09:26:08 +0000 (11:26 +0200)]
Updated SWID scenarios
Andreas Steffen [Mon, 14 Apr 2014 09:25:41 +0000 (11:25 +0200)]
swid_generator software-id does not generate empty lines any more
Andreas Steffen [Sat, 12 Apr 2014 08:56:16 +0000 (10:56 +0200)]
Added result information to TPMRA workitems
On the occasion got rid of complicated functional component stuff
Andreas Steffen [Fri, 11 Apr 2014 14:58:08 +0000 (16:58 +0200)]
Indicate IMV in assessment log statement
Andreas Steffen [Fri, 11 Apr 2014 14:30:55 +0000 (16:30 +0200)]
Implemented segmented SWID tag attributes on IMV side
Andreas Steffen [Thu, 10 Apr 2014 08:25:39 +0000 (10:25 +0200)]
Use python-based swidGenerator to generated SWID tags
Andreas Steffen [Mon, 7 Apr 2014 13:17:32 +0000 (15:17 +0200)]
Updated imv database templates
Andreas Steffen [Sun, 6 Apr 2014 05:18:28 +0000 (07:18 +0200)]
Optimized PTS measurements
Andreas Steffen [Sat, 5 Apr 2014 14:11:13 +0000 (16:11 +0200)]
Use cached pid for product-based package access
Andreas Steffen [Sat, 5 Apr 2014 13:38:06 +0000 (15:38 +0200)]
Make Attestation IMV independent of OS IMV
Andreas Steffen [Fri, 4 Apr 2014 21:00:40 +0000 (23:00 +0200)]
Separated IMV session management from IMV policy database
Andreas Steffen [Wed, 2 Apr 2014 04:54:16 +0000 (06:54 +0200)]
Renamed the AIK public key parameter to imc-attestation.aik_pubkey
Andreas Steffen [Mon, 31 Mar 2014 11:00:40 +0000 (13:00 +0200)]
Implemented configurable Device ID in OS IMC
Andreas Steffen [Tue, 15 Apr 2014 07:20:38 +0000 (09:20 +0200)]
Version bump to 5.2.0dr1
Andreas Steffen [Mon, 14 Apr 2014 13:18:38 +0000 (15:18 +0200)]
Version bump to 5.1.3
Tobias Brunner [Mon, 14 Apr 2014 11:32:36 +0000 (13:32 +0200)]
NEWS: Added info about CVE-2014-2338
Martin Willi [Thu, 20 Feb 2014 15:08:43 +0000 (16:08 +0100)]
ikev2: Reject CREATE_CHILD_SA exchange on unestablished IKE_SAs
Prevents a responder peer to trick us into established state by starting
IKE_SA rekeying before the IKE_SA has been authenticated during IKE_AUTH.
Fixes CVE-2014-2338.
Tobias Brunner [Wed, 9 Apr 2014 16:04:33 +0000 (18:04 +0200)]
eap-mschapv2: Fix potential leaks in case of invalid messages from servers
Tobias Brunner [Wed, 9 Apr 2014 15:47:32 +0000 (17:47 +0200)]
pts: Make sure the complete AIK blob has been read
Tobias Brunner [Wed, 9 Apr 2014 15:09:55 +0000 (17:09 +0200)]
attr: Don't shift the 32-bit netmask by 32
This is undefined behavior as per the C99 standard (sentence 1185):
"If the value of the right operand is negative or is greater or equal
to the width of the promoted left operand, the behavior is undefined."
Apparently shifts may be done modulo the width on some platforms so
a shift by 32 would not shift at all.
Tobias Brunner [Wed, 9 Apr 2014 14:35:46 +0000 (16:35 +0200)]
nm: Fix NULL-pointer dereference when handling TUN device failure
Tobias Brunner [Wed, 9 Apr 2014 13:28:54 +0000 (15:28 +0200)]
x509: Don't include authKeyIdentifier in self-signed certificates
As the comment indicates this was the intention in
d7be2906433a7dcfefc1fd732587865688dbfe1b all along.
Tobias Brunner [Wed, 9 Apr 2014 13:18:13 +0000 (15:18 +0200)]
x509: Initialize certs when building optionalSignature for OCSP requests
Tobias Brunner [Wed, 9 Apr 2014 14:05:55 +0000 (16:05 +0200)]
stroke: Fix memory leak when printing unknown AC group OIDs
Tobias Brunner [Wed, 9 Apr 2014 13:53:35 +0000 (15:53 +0200)]
pki: Fix memory leak when printing unknown AC group OIDs
Tobias Brunner [Wed, 9 Apr 2014 13:12:27 +0000 (15:12 +0200)]
pki: Removed extra continue statement
Andreas Steffen [Tue, 8 Apr 2014 11:09:03 +0000 (13:09 +0200)]
Added support for msSmartcardLogon EKU
Andreas Steffen [Tue, 8 Apr 2014 09:32:30 +0000 (11:32 +0200)]
Added some more OIDs
Andreas Steffen [Mon, 7 Apr 2014 11:29:39 +0000 (13:29 +0200)]
Initialize m1 to suppress compiler warning
Andreas Steffen [Sun, 6 Apr 2014 15:54:55 +0000 (17:54 +0200)]
Fixed another dirname/basename refactoring bug.
file was freed before use.
Andreas Steffen [Sun, 6 Apr 2014 15:26:52 +0000 (17:26 +0200)]
Fixed dirname/basename refactoring bug.
Variables used in a database query have to be kept until the end of the enumeration
Andreas Steffen [Fri, 4 Apr 2014 21:44:55 +0000 (23:44 +0200)]
Added SHA3 OIDs
Andreas Steffen [Fri, 4 Apr 2014 21:04:54 +0000 (23:04 +0200)]
Fixed pretest script in tnc/tnccs-20-pt-tls scenario
Tobias Brunner [Wed, 2 Apr 2014 13:06:56 +0000 (15:06 +0200)]
ike-cfg: Properly compare IKE proposals for equality
Tobias Brunner [Tue, 1 Apr 2014 08:30:13 +0000 (10:30 +0200)]
leak-detective: LEAK_DETECTIVE_DISABLE completely disables LD
If lib->leak_detective is non-null some code parts (e.g. the plugin
loader) assume LD is actually used.
Tobias Brunner [Tue, 2 Jul 2013 12:01:38 +0000 (14:01 +0200)]
testing: Run 'conntrack -F' before all test scenarios
This prevents failures due to remaining conntrack entries.
Tobias Brunner [Wed, 2 Apr 2014 09:50:11 +0000 (11:50 +0200)]
unit-tests: Verify two bytes at once when testing chunk_clear()
This reduces the chances of arbitrary test failures if the memory area
already got overwritten.
Martin Willi [Tue, 1 Apr 2014 12:53:28 +0000 (14:53 +0200)]
Merge branch 'tls-unit-tests'
Add some initial unit-tests to libtls, testing all supported cipher suites
against self, both with and without client authentication, for all supported
TLS versions.
Martin Willi [Tue, 25 Mar 2014 08:25:14 +0000 (09:25 +0100)]
tls: Add a test case to check correct enum name mapping of cipher suites
Martin Willi [Mon, 24 Mar 2014 17:01:00 +0000 (18:01 +0100)]
tls: Add socket based tests testing all supported suites with TLS 1.2/1.1/1.0
Martin Willi [Tue, 1 Apr 2014 09:45:45 +0000 (11:45 +0200)]
tls: Remove superfluous initializers in TLS AEAD implementations
Martin Willi [Tue, 25 Mar 2014 09:12:51 +0000 (10:12 +0100)]
tls: Support a maximum TLS version to negotiate using TLS socket abstraction
Martin Willi [Tue, 25 Mar 2014 09:19:41 +0000 (10:19 +0100)]
tls: Support a null encryption flag on TLS socket abstraction
Martin Willi [Tue, 25 Mar 2014 08:49:04 +0000 (09:49 +0100)]
tls: Introduce a generic TLS purpose that accepts NULL encryption ciphers
Martin Willi [Mon, 24 Mar 2014 13:28:50 +0000 (14:28 +0100)]
tls: Export a function to list supported TLS cipher suites
Martin Willi [Mon, 24 Mar 2014 12:47:03 +0000 (13:47 +0100)]
tls: Create a unit-test runner
Martin Willi [Tue, 25 Mar 2014 13:14:37 +0000 (14:14 +0100)]
unit-tests: Catch timeouts during test runner deinit function
The test runner deinit function often cancels all threads from the pool. This
operation might hang on error conditions, hence we should include that hook in
the test timeout to fail properly.
Martin Willi [Mon, 24 Mar 2014 16:17:50 +0000 (17:17 +0100)]
unit-tests: Prevent a failing worker thread to go wild after it fails
A worker raises SIGUSR1 to inform the main thread that the test fails. The main
thread then starts cancelling workers, but the offending thread should be
terminated immediately to prevent it from test continuation.
Andreas Steffen [Tue, 1 Apr 2014 08:12:15 +0000 (10:12 +0200)]
Test TLS AEAD cipher suites
Andreas Steffen [Mon, 31 Mar 2014 20:22:51 +0000 (22:22 +0200)]
Added Ubuntu 14.04 to IMV database
Andreas Steffen [Mon, 31 Mar 2014 20:08:50 +0000 (22:08 +0200)]
Slightly edited evaltest of ikev2/ocsp-untrusted-cert scenario
Tobias Brunner [Mon, 31 Mar 2014 14:58:46 +0000 (16:58 +0200)]
unit-tests: Always load address of testable functions
The addresses can actually change as plugins are loaded/unloaded for
each test case.
Fixes #551.
Tobias Brunner [Mon, 31 Mar 2014 12:47:00 +0000 (14:47 +0200)]
settings: Reduce log verbosity if strongswan.conf does not exist
In some situations we expect strongswan.conf to not exist, for instance,
when running the unit tests before installation.
Tobias Brunner [Thu, 13 Mar 2014 13:05:29 +0000 (14:05 +0100)]
test-vectors: Renumber AES-GCM test vectors according to original source
Also adds several missing ones.
Martin Willi [Mon, 31 Mar 2014 14:17:57 +0000 (16:17 +0200)]
Merge branch 'tls-aead'
Adds AEAD support to the TLS stack, currently supporting AES-GCM. Brings fixes
for TLS record fragmentation, enforcing TLS versions < 1.2 and proper signature
scheme support indication.
Martin Willi [Mon, 31 Mar 2014 14:17:15 +0000 (16:17 +0200)]
NEWS: Note TLS AEAD mode
Martin Willi [Tue, 25 Mar 2014 08:20:15 +0000 (09:20 +0100)]
tls: Fix some TLS cipher suite enum names
It is important to have them mapped correctly, as we use these official TLS
identifiers to configure specific TLS suites.
Martin Willi [Tue, 25 Mar 2014 09:50:51 +0000 (10:50 +0100)]
tls: Include TLS version announced in Client Hello in encrypted premaster
While a hardcoded 1.2 version is fine when we offer that in Client Hello, we
should include the actually offered version if it has been reduced before
starting the exchange.
Martin Willi [Fri, 21 Mar 2014 08:29:44 +0000 (09:29 +0100)]
tls: Check for minimal TLS record length before each record iteration
Fixes fragment reassembling if a buffer contains more than one record, but
the last record contains a partial TLS record header. Thanks to Nick Saunders
and Jamil Nimeh for identifying this issue and providing a fix for it.
Martin Willi [Tue, 11 Mar 2014 09:57:18 +0000 (10:57 +0100)]
tls: Fix AEAD algorithm filtering, avoid filtering all suites if no AEAD found
Martin Willi [Wed, 15 Jan 2014 14:51:03 +0000 (15:51 +0100)]
tls: Offer TLS signature schemes in ClientHello in order of preference
Additionally, we now query plugin features to find out what schemes we exactly
support.
Martin Willi [Mon, 3 Feb 2014 17:08:11 +0000 (18:08 +0100)]
tls: Define AES-GCM cipher suites from RFC 5288/5289
Martin Willi [Mon, 3 Feb 2014 17:03:41 +0000 (18:03 +0100)]
tls: Implement the TLS AEAD abstraction for real AEAD modes
Martin Willi [Mon, 3 Feb 2014 12:20:46 +0000 (13:20 +0100)]
tls: Separate TLS protection to abstracted AEAD modes
To better separate the code path for different TLS versions and modes of
operation, we introduce a TLS AEAD abstraction. We provide three implementations
using traditional transforms, and get prepared for TLS AEAD modes.
Martin Willi [Fri, 31 Jan 2014 14:53:38 +0000 (15:53 +0100)]
aead: Support custom AEAD salt sizes
The salt, or often called implicit nonce, varies between AEAD algorithms and
their use in protocols. For IKE and ESP, GCM uses 4 bytes, while CCM uses
3 bytes. With TLS, however, AEAD mode uses 4 bytes for both GCM and CCM.
Our GCM backends currently support 4 bytes and CCM 3 bytes only. This is fine
until we go for CCM mode support in TLS, which requires 4 byte nonces.
Martin Willi [Thu, 27 Feb 2014 08:36:46 +0000 (09:36 +0100)]
ikev2: Recreate a CHILD_SA that got a hard lifetime expire without rekeying
Works around issues related to system time changes and kernel backends using
that system time, such as Linux XFRM.
Martin Willi [Mon, 31 Mar 2014 12:53:15 +0000 (14:53 +0200)]
revocation: Log error if no OCSP signer candidate found
Fixes evaluation of ikev2/ocsp-untrusted-cert.
Martin Willi [Mon, 31 Mar 2014 12:44:50 +0000 (14:44 +0200)]
Merge branch 'ocsp-constraints'
Limits cached OCSP verification to responses signed by the CA, a directly
delegated signer or a pre-installed OCSP responder certificate. Disables
auth config merge for revocation trust-chain strength checkin, as it breaks
CA constraints in some scenarios.
Martin Willi [Tue, 25 Mar 2014 13:34:58 +0000 (14:34 +0100)]
revocation: Restrict OCSP signing to specific certificates
To avoid considering each cached OCSP response and evaluating its trustchain,
we limit the certificates considered for OCSP signing to:
- The issuing CA of the checked certificate
- A directly delegated signer by the same CA, having the OCSP signer constraint
- Any locally installed (trusted) certificate having the OCSP signer constraint
The first two options cover the requirements from RFC 6960 2.6. For
compatibility with non-conforming CAs, we allow the third option as exception,
but require the installation of such certificates locally.
Martin Willi [Thu, 27 Mar 2014 09:59:29 +0000 (10:59 +0100)]
revocation: Don't merge auth config of CLR/OCSP trustchain validation
This behavior was introduced with
6840a6fb to avoid key/signature strength
checking for the revocation trustchain as we do it for end entity certificates.
Unfortunately this breaks CA constraint checking under certain conditions, as
we merge additional intermediate/CA certificates to the auth config.
As key/signature strength checking of the revocation trustchain is a rather
exotic requirement we drop support for that to properly enforce CA constraints.