Martin Willi [Tue, 15 Jan 2013 16:38:10 +0000 (17:38 +0100)]
Add a libpttls providing NEA PT-TLS / TNC IF-T for TLS transport layer
Martin Willi [Tue, 15 Jan 2013 16:36:35 +0000 (17:36 +0100)]
Send TLS close notify during tls_socket_t destruction
Martin Willi [Tue, 15 Jan 2013 16:36:11 +0000 (17:36 +0100)]
Send TLS close notify if application returns SUCCESS
Martin Willi [Mon, 14 Jan 2013 14:32:12 +0000 (15:32 +0100)]
Block TLS read when sending data, but have to wait for the handshake data first
Martin Willi [Mon, 14 Jan 2013 12:04:51 +0000 (13:04 +0100)]
TNCCS plugins don't depend on EAP-TNC, but can be used by other transports, too
Martin Willi [Fri, 11 Jan 2013 14:27:24 +0000 (15:27 +0100)]
Add a bio_reader_t constructor variant freeing passed data during destruction
Martin Willi [Thu, 10 Jan 2013 15:20:06 +0000 (16:20 +0100)]
Use a more POSIXy tls_socket interface with more flexibility.
If an unsufficient read buffer is provided, application data gets cached
for subsequent read() calls.
Martin Willi [Thu, 3 Jan 2013 13:09:05 +0000 (14:09 +0100)]
Add a chunk_from_str() initializer that does not include 0-terminator
Martin Willi [Tue, 15 Jan 2013 12:27:35 +0000 (13:27 +0100)]
Reseed rdrand after every 128bit sample only
Andreas Steffen [Tue, 15 Jan 2013 01:41:22 +0000 (02:41 +0100)]
version bump to 5.0.2rc1
Tobias Brunner [Mon, 14 Jan 2013 16:23:52 +0000 (17:23 +0100)]
android: Properly escape apostrophes in Ukrainian translation
Tobias Brunner [Mon, 14 Jan 2013 16:21:54 +0000 (17:21 +0100)]
android: Implement kernel_net_t.get_interface via JNI
This is now required to properly accept/install a virtual IP address.
Fixes #275.
Tobias Brunner [Mon, 14 Jan 2013 16:18:35 +0000 (17:18 +0100)]
android: Moved chunk_from_byte_array and byte_array_from_chunk helper functions
Tobias Brunner [Mon, 14 Jan 2013 16:16:18 +0000 (17:16 +0100)]
android: Set OPENSSL_NO_CMS in Android.mk as it is not set in opensslconf.h on Android
Tobias Brunner [Mon, 14 Jan 2013 11:11:24 +0000 (12:11 +0100)]
Properly send IKEv1 packets if no ike_cfg is known yet
This applies for error notifies.
Martin Willi [Mon, 14 Jan 2013 09:33:14 +0000 (10:33 +0100)]
Don't handle right=%any6 as "loose" identity, but as %any
Martin Willi [Mon, 14 Jan 2013 09:26:12 +0000 (10:26 +0100)]
Respect given address family when resolving "%any"
Tobias Brunner [Mon, 14 Jan 2013 08:16:33 +0000 (09:16 +0100)]
Android.mk of libstrongswan updated
Tobias Brunner [Sat, 12 Jan 2013 10:58:26 +0000 (11:58 +0100)]
Merge branch 'ikev1-fragmentation'
This adds support for the proprietary IKEv1 fragmentation extension.
Conflicts:
NEWS
Tobias Brunner [Sat, 12 Jan 2013 10:51:35 +0000 (11:51 +0100)]
NEWS for fragmentation extension added
Conflicts:
NEWS
Tobias Brunner [Sat, 12 Jan 2013 10:48:32 +0000 (11:48 +0100)]
Added an option to configure the maximum size of a fragment
Tobias Brunner [Sat, 12 Jan 2013 10:39:03 +0000 (11:39 +0100)]
Properly detect fragmentation capabilities
Cisco sends 0xc0000000 so we check that part of the VID separately.
Tobias Brunner [Mon, 24 Dec 2012 11:59:30 +0000 (12:59 +0100)]
Added an option that allows to force IKEv1 fragmentation
Martin Willi [Fri, 11 Jan 2013 13:45:32 +0000 (14:45 +0100)]
Don't use bio_writer_t.skip() to write length field when appending more data
If the writer reallocates its buffer, the length pointer might not be valid
anymore, or even worse, point to an arbitrary allocation.
Martin Willi [Fri, 11 Jan 2013 09:48:57 +0000 (10:48 +0100)]
Add rdrand NEWS
Martin Willi [Fri, 4 Jan 2013 15:34:56 +0000 (16:34 +0100)]
Use raw opcodes for rdrand to build with older binutils
Martin Willi [Fri, 4 Jan 2013 16:34:07 +0000 (17:34 +0100)]
Provide RNG_TRUE quality in rdrand by mixing reseeded outputs using AES
Martin Willi [Fri, 4 Jan 2013 15:07:31 +0000 (16:07 +0100)]
Provide RNG_STRONG quality in rdrand by forcing PRNG reseed after every sample
Martin Willi [Fri, 4 Jan 2013 14:33:10 +0000 (15:33 +0100)]
Provide RNG_WEAK quality random generator in rdrand
Martin Willi [Fri, 4 Jan 2013 13:33:45 +0000 (14:33 +0100)]
Add a rdrand plugin stub detecting availability of RDRAND instructions
Martin Willi [Fri, 11 Jan 2013 09:31:25 +0000 (10:31 +0100)]
Add NEWS about improved Windows IKEv1 compatibility
Martin Willi [Fri, 11 Jan 2013 09:24:23 +0000 (10:24 +0100)]
Streamline debug output when receiving intermediate CA certificates in IKEv1
Martin Willi [Fri, 11 Jan 2013 09:20:06 +0000 (10:20 +0100)]
Refactored IKEv2 cert/certreq payload processing to multiple functions
Martin Willi [Fri, 11 Jan 2013 09:19:12 +0000 (10:19 +0100)]
Refactored IKEv1 cert payload processing to multiple functions
Volker Rümelin [Thu, 10 Jan 2013 20:27:20 +0000 (21:27 +0100)]
IKEv1 support for PKCS#7 wrapped certificates
Volker Rümelin [Thu, 10 Jan 2013 20:24:37 +0000 (21:24 +0100)]
Fixed some typos in comments
Pavel Kopchyk [Wed, 9 Jan 2013 04:30:55 +0000 (05:30 +0100)]
Fixed some typos in Ukrainian translation
Thomas Klute [Fri, 4 Jan 2013 12:47:15 +0000 (13:47 +0100)]
conftest: Add support for time_format and ike_name options in log sections
Both options are well supported for normal operation but were completely
ignored by conftest, which used hard coded defaults. File options are
still missing but could be added in a similar way.
Thomas Klute [Thu, 3 Jan 2013 15:03:44 +0000 (16:03 +0100)]
conftest: Fix log level settings for stdout
This patch fixes bug #272 ("conftest ignores log settings for stdout").
http://wiki.strongswan.org/issues/272
According to the documentation of add_logger in src/libcharon/bus/bus.h,
the relevant log levels of a logger are registered with the logging
subsystem when adding the logger. If the log levels change later, the
logger must be re-added to propagate the new settings. In conftest.c,
the stdout logger is initialized and added before reading the logging
settings, but wasn't re-added after reading the settings.
Thomas Klute [Wed, 19 Dec 2012 13:14:55 +0000 (14:14 +0100)]
conftest: Make outgoing sequence number set by reset_seq configurable
This is useful for certain test cases. Passing the sequence number to
the callback requires a new struct that contains both the number and the
xfrm_usersa_id. The new configuration parameter is called oseq in
accordance with the kernel name, see the comment in the reset_cb
callback function for details.
Martin Willi [Thu, 3 Jan 2013 10:12:05 +0000 (11:12 +0100)]
Include opensslconf.h before checking its defines
Martin Willi [Thu, 3 Jan 2013 10:05:49 +0000 (11:05 +0100)]
Don't build OpenSSL PKCS#7 code if OPENSSL_NO_CMS defined
Andreas Steffen [Wed, 26 Dec 2012 08:28:17 +0000 (09:28 +0100)]
make pacman.sh run under cron
Tobias Brunner [Mon, 24 Dec 2012 11:28:01 +0000 (12:28 +0100)]
Use a connection specific option to en-/disable IKEv1 fragmentation
Tobias Brunner [Fri, 21 Dec 2012 17:40:23 +0000 (18:40 +0100)]
Include source port in init hash for fragmented messages
Tobias Brunner [Fri, 21 Dec 2012 17:27:02 +0000 (18:27 +0100)]
Add an option to en-/disable IKE fragmentation
Fragments are always accepted but will not be sent if disabled. The
vendor ID is only sent if the option is enabled.
Tobias Brunner [Fri, 21 Dec 2012 17:25:06 +0000 (18:25 +0100)]
Split larger messages into fragments if IKE fragmentation is supported by peer
Tobias Brunner [Fri, 21 Dec 2012 16:49:45 +0000 (17:49 +0100)]
Log message size for in- and outbound IKE messages
Tobias Brunner [Thu, 20 Dec 2012 15:04:38 +0000 (16:04 +0100)]
Add support to create IKE fragments
All fragments currently use the same fragment ID (1) as that's what
other implementations are doing.
Tobias Brunner [Wed, 19 Dec 2012 13:07:06 +0000 (14:07 +0100)]
Log added NAT-T vendor IDs
Tobias Brunner [Thu, 20 Dec 2012 11:16:08 +0000 (12:16 +0100)]
Detect a peer's support for IKE fragmentation
Fragments are accepted even if this vendor ID is not seen.
Tobias Brunner [Thu, 20 Dec 2012 11:14:25 +0000 (12:14 +0100)]
Map fragmented initial initial Main or Aggressive Mode messages to the same IKE_SA
Tobias Brunner [Thu, 20 Dec 2012 10:55:33 +0000 (11:55 +0100)]
Allow ID_PROT/AGGRESSIVE messages for established IKE_SAs if they contain fragments
Other implementations send fragments always in an initial message type
even for transaction or quick mode exchanges.
Tobias Brunner [Wed, 12 Dec 2012 17:29:31 +0000 (18:29 +0100)]
Don't handle fragmented messages larger than charon.max_packet
Tobias Brunner [Wed, 12 Dec 2012 17:23:11 +0000 (18:23 +0100)]
Don't update an IKE_SA-entry's cached message ID when handling fragments
Tobias Brunner [Wed, 12 Dec 2012 17:22:32 +0000 (18:22 +0100)]
Store inbound IKE fragments and reassemble the message when all fragments are received
Andreas Steffen [Mon, 24 Dec 2012 11:06:07 +0000 (12:06 +0100)]
deleted newly constructed attributes in send_assessment
Dmitry Korzhevin [Mon, 24 Dec 2012 10:44:28 +0000 (11:44 +0100)]
Added Russian and Ukrainian strings for Android client
Tobias Brunner [Fri, 21 Dec 2012 15:40:58 +0000 (16:40 +0100)]
Add message rules to properly handle IKE fragments
These are sent in unencrypted messages and are the only payload
contained in such messages.
Tobias Brunner [Wed, 12 Dec 2012 17:18:37 +0000 (18:18 +0100)]
Reset the encrypted flag when handling IKE messages that contain a fragment
Racoon sets the encrypted bit for messages containing a fragment, but these
messages are not really encrypted (the fragmented message is though).
Tobias Brunner [Wed, 12 Dec 2012 17:16:58 +0000 (18:16 +0100)]
Payload added to handle IKE fragments
Martin Willi [Fri, 21 Dec 2012 08:48:35 +0000 (09:48 +0100)]
Add parantheses to avoid compiler warning
Andreas Steffen [Sun, 23 Dec 2012 21:16:30 +0000 (22:16 +0100)]
Send empty CDATA batch if TNC client has no data to send
Tobias Brunner [Thu, 20 Dec 2012 08:31:38 +0000 (09:31 +0100)]
Fixed some typos, courtesy of codespell
Adrian-Ken Rueegsegger [Wed, 19 Dec 2012 14:48:35 +0000 (15:48 +0100)]
Raise an alert if IKE SA is kept
This alert is raised when the establishment of a child SA fails but the
IKE SA is kept.
Reto Buerki [Tue, 18 Dec 2012 15:11:19 +0000 (16:11 +0100)]
stroke: Drop unneeded [MY|OTHER]_NETBITS
Reto Buerki [Wed, 18 Jul 2012 14:19:31 +0000 (16:19 +0200)]
stroke: Enable install_policy in add_connection()
Volker Rümelin [Sat, 15 Dec 2012 13:11:26 +0000 (14:11 +0100)]
Add support for draft-ietf-ipsec-nat-t-ike-03 and earlier
This adds support for early versions of the draft that eventually
resulted in RFC 3947.
Martin Willi [Wed, 19 Dec 2012 09:43:35 +0000 (10:43 +0100)]
NEWS about error-notify
Martin Willi [Wed, 21 Nov 2012 10:12:53 +0000 (11:12 +0100)]
Add missing error_notify_msg.h to distribution tarball
Martin Willi [Thu, 8 Nov 2012 10:22:26 +0000 (11:22 +0100)]
Add an error-notify sample application to listen to error notifications
Martin Willi [Tue, 6 Nov 2012 15:46:49 +0000 (16:46 +0100)]
Add an error-notify plugin to send catched alerts to listening applications
Martin Willi [Tue, 6 Nov 2012 14:26:15 +0000 (15:26 +0100)]
Raise an alert if half-open timeout limit reached
Martin Willi [Tue, 6 Nov 2012 10:48:58 +0000 (11:48 +0100)]
Raise an alert if an authorize() hook fails
Martin Willi [Tue, 6 Nov 2012 10:43:19 +0000 (11:43 +0100)]
Raise an alert if allocating virtual IPs fails
Martin Willi [Tue, 6 Nov 2012 10:33:29 +0000 (11:33 +0100)]
Raise an alert if kernel policy installation fails
Martin Willi [Tue, 6 Nov 2012 10:32:18 +0000 (11:32 +0100)]
Raise an alert if kernel SA installation fails
Martin Willi [Tue, 6 Nov 2012 10:27:38 +0000 (11:27 +0100)]
Raise an alert on traffic selector mismatch
Martin Willi [Tue, 6 Nov 2012 10:19:52 +0000 (11:19 +0100)]
Raise alerts when enforcing IKE_SA unique policy
Martin Willi [Tue, 6 Nov 2012 10:05:04 +0000 (11:05 +0100)]
Raise an alert if CHILD_SA proposals mismatch
Martin Willi [Tue, 6 Nov 2012 10:01:49 +0000 (11:01 +0100)]
Raise an alert if IKE proposals mismatch
Martin Willi [Mon, 5 Nov 2012 14:33:34 +0000 (15:33 +0100)]
Raise an alert of generating local authentication data fails
Martin Willi [Wed, 19 Dec 2012 09:38:42 +0000 (10:38 +0100)]
Add NEWS about BER capable OpenSSL PKCS#7 backend
Martin Willi [Thu, 29 Nov 2012 15:48:18 +0000 (16:48 +0100)]
Free leaking scep attributes
Martin Willi [Thu, 29 Nov 2012 15:35:41 +0000 (16:35 +0100)]
Corrected error message if enveloped-data decryption fails
Martin Willi [Thu, 29 Nov 2012 15:35:06 +0000 (16:35 +0100)]
Fix up serialNumber in openssl PKCS#7 if it has a leading MSB set
Martin Willi [Thu, 29 Nov 2012 15:12:45 +0000 (16:12 +0100)]
Don't handle PKCS#7 containers with infinite length encodings in pkcs7 plugin
Martin Willi [Thu, 29 Nov 2012 14:56:53 +0000 (15:56 +0100)]
Implement PKCS#7 decryption using openssl
Martin Willi [Thu, 29 Nov 2012 13:39:35 +0000 (14:39 +0100)]
Make available wrapped certificates while verifying PKCS#7 signatures in openssl
Martin Willi [Thu, 29 Nov 2012 13:30:08 +0000 (14:30 +0100)]
Implement openssl PKCS#7 certficiate enumeration
Martin Willi [Thu, 29 Nov 2012 11:02:07 +0000 (12:02 +0100)]
Fix doxygen grouping regarding containers and PKCS#7
Martin Willi [Thu, 29 Nov 2012 10:53:13 +0000 (11:53 +0100)]
Enable pkcs7 plugin when building scepclient on Android
Martin Willi [Thu, 29 Nov 2012 10:52:27 +0000 (11:52 +0100)]
Move PKCS#9 attribute lists to pkcs7 plugin, as we currently use it there only
Martin Willi [Thu, 29 Nov 2012 10:39:49 +0000 (11:39 +0100)]
Implement get_attribute() in openssl PKCS#7 backend
Martin Willi [Thu, 29 Nov 2012 10:29:46 +0000 (11:29 +0100)]
Allocate data returned by pkcs7_t.get_attribute()
Martin Willi [Wed, 28 Nov 2012 17:45:30 +0000 (18:45 +0100)]
Implement OpenSSL PKCS#7 signed-data parsing and verification
Martin Willi [Wed, 28 Nov 2012 13:59:49 +0000 (14:59 +0100)]
Add a stub for OpenSSL PKCS#7 parsing
Martin Willi [Wed, 28 Nov 2012 11:44:05 +0000 (12:44 +0100)]
Remove unused monolithic PKCS#7 code
Martin Willi [Wed, 28 Nov 2012 11:41:38 +0000 (12:41 +0100)]
Migrated scepclient to new modular PKCS# API
Martin Willi [Wed, 28 Nov 2012 11:41:15 +0000 (12:41 +0100)]
Fix encryption algorithm/key size argument processing in PKCS#7 enveloped-data