3 years agosmp: Use correct printf specifier to print SPIs
Tobias Brunner [Tue, 7 May 2019 12:50:11 +0000 (14:50 +0200)]
smp: Use correct printf specifier to print SPIs

3 years agofast: Use correct printf specifier to print content length
Tobias Brunner [Tue, 7 May 2019 12:48:19 +0000 (14:48 +0200)]
fast: Use correct printf specifier to print content length

3 years agolibimcv: Use proper printf specifier for unsigned issuer and responder IDs
Tobias Brunner [Tue, 7 May 2019 12:43:30 +0000 (14:43 +0200)]
libimcv: Use proper printf specifier for unsigned issuer and responder IDs

3 years agoswima-collector: Use proper type for field precision
Tobias Brunner [Tue, 7 May 2019 12:40:58 +0000 (14:40 +0200)]
swima-collector: Use proper type for field precision

3 years agoopenssl: Fix build with OpenSSL 1.1.1 without compatibility layer
Tobias Brunner [Tue, 7 May 2019 09:44:34 +0000 (11:44 +0200)]
openssl: Fix build with OpenSSL 1.1.1 without compatibility layer

If OpenSSL is built with --api, defines for deprecated functions in
OpenSSL's header files are not visible anymore.

Fixes #3045.

3 years agotravis: Build OpenSSL 1.1.1 without compatibility layer for older versions
Tobias Brunner [Tue, 7 May 2019 09:28:29 +0000 (11:28 +0200)]
travis: Build OpenSSL 1.1.1 without compatibility layer for older versions

Configuring 1.1.1 is not actually possible with 1.1.1b, not sure if
that's on purpose.

3 years agotravis: Make sure crypto plugins are actually loaded
Tobias Brunner [Tue, 7 May 2019 11:43:45 +0000 (13:43 +0200)]
travis: Make sure crypto plugins are actually loaded

3 years agostarter: Remove IPsec stack detection
Tobias Brunner [Tue, 7 May 2019 09:03:23 +0000 (11:03 +0200)]
starter: Remove IPsec stack detection

Checking specifically for /proc/net/pfkey is not ideal as af_key will
eventually be removed in Linux kernels.  Support for KLIPS is long gone.
The detection also wasn't used for anything anymore (failures were just
ignored since the ports to BSD-based systems).  And modprobing doesn't seem
to be necessary either (charon-systemd doesn't do that, for instance).

3 years agovici: Add Python command wrappers to tarball
Tobias Brunner [Mon, 6 May 2019 13:51:05 +0000 (15:51 +0200)]
vici: Add Python command wrappers to tarball

Fixes: e0f7da864481 ("vici: Extract command wrappers in Python bindings")

3 years agopki: Fix memory leaks in --signcrl if signature scheme is not found
Tobias Brunner [Tue, 30 Apr 2019 08:25:56 +0000 (10:25 +0200)]
pki: Fix memory leaks in --signcrl if signature scheme is not found

Fixes: dd4bd21c5a22 ("pki: Query private key for supported signature schemes")

3 years agotesting: Update documentation in headers of all updown scripts
Tobias Brunner [Mon, 29 Apr 2019 15:43:04 +0000 (17:43 +0200)]
testing: Update documentation in headers of all updown scripts

3 years agoswanctl: Move documentation of if_id_in/out after all mark-related options
Tobias Brunner [Mon, 29 Apr 2019 15:37:30 +0000 (17:37 +0200)]
swanctl: Move documentation of if_id_in/out after all mark-related options

Also fix a typo.

3 years agoFixed some typos, courtesy of codespell
Tobias Brunner [Mon, 29 Apr 2019 13:07:25 +0000 (15:07 +0200)]
Fixed some typos, courtesy of codespell

3 years agononce: Allow overriding the RNG quality used to generate nonces
Tobias Brunner [Tue, 23 Apr 2019 09:14:44 +0000 (11:14 +0200)]
nonce: Allow overriding the RNG quality used to generate nonces

Usually, changing this won't be necessary (actually, some plugins
specifically use different DRGBs for RNG_WEAK in order to separate
the public nonces from random data used for e.g. DH).
But for experts with special plugin configurations this might be
more flexible and avoids code changes.

3 years agounit-tests: Fix skipping of some ECDSA signature schemes
SophieK [Mon, 29 Apr 2019 06:33:47 +0000 (14:33 +0800)]
unit-tests: Fix skipping of some ECDSA signature schemes

Closes strongswan/strongswan#137.

3 years agoNEWS: Added some news for 5.8.0
Tobias Brunner [Fri, 26 Apr 2019 16:54:58 +0000 (18:54 +0200)]
NEWS: Added some news for 5.8.0

3 years agoMerge branch 'update-vici-bindings'
Tobias Brunner [Fri, 26 Apr 2019 08:19:21 +0000 (10:19 +0200)]
Merge branch 'update-vici-bindings'

Updates the command wrappers in all the bindings and simplifies calling
new commands (i.e. not yet wrapped) with the Python and Ruby bindings.

Fixes #3028.

3 years agovici: Update command wrappers in the Perl bindings
Tobias Brunner [Thu, 25 Apr 2019 09:09:20 +0000 (11:09 +0200)]
vici: Update command wrappers in the Perl bindings

Note that load_key() now returns the complete response (to get the key

3 years agovici: Update some data in the Ruby gemspec
Tobias Brunner [Thu, 25 Apr 2019 08:26:11 +0000 (10:26 +0200)]
vici: Update some data in the Ruby gemspec

3 years agovici: Some code style fixes in the Ruby bindings
Tobias Brunner [Wed, 24 Apr 2019 16:05:11 +0000 (18:05 +0200)]
vici: Some code style fixes in the Ruby bindings

As reported by rubocop (some issues were not fixed, in particular
related to class/method length metrics).

3 years agovici: Update command wrappers of the Ruby bindings
Tobias Brunner [Wed, 24 Apr 2019 14:37:11 +0000 (16:37 +0200)]
vici: Update command wrappers of the Ruby bindings

Also reorder them to match

3 years agovici: Refactor how commands are called in the Ruby bindings
Tobias Brunner [Wed, 24 Apr 2019 14:05:12 +0000 (16:05 +0200)]
vici: Refactor how commands are called in the Ruby bindings

Also expose a method to call arbitrary commands, which allows calling not
yet wrapped commands. Exceptions are raised for all commands if the response
includes a negative "success" key (similar to how it's done in the Python

3 years agovici: Fix formatting of return values for load-conn and load-authority commands
Tobias Brunner [Tue, 23 Apr 2019 17:56:22 +0000 (19:56 +0200)]
vici: Fix formatting of return values for load-conn and load-authority commands

3 years agovici: Add missing command wrappers for Python bindings
Tobias Brunner [Tue, 23 Apr 2019 14:13:19 +0000 (16:13 +0200)]
vici: Add missing command wrappers for Python bindings

Also change some for which the return value became relevant.

3 years agovici: Extract command wrappers in Python bindings
Tobias Brunner [Thu, 18 Apr 2019 08:56:15 +0000 (10:56 +0200)]
vici: Extract command wrappers in Python bindings

This simplifies the interface and allows calling not yet wrapped
commands more easily.

3 years agoeap-aka-3gpp2: Increase SQN after each authentication
Tobias Brunner [Wed, 24 Apr 2019 09:30:14 +0000 (11:30 +0200)]
eap-aka-3gpp2: Increase SQN after each authentication

3 years agoMerge branch 'childless'
Tobias Brunner [Thu, 25 Apr 2019 13:32:02 +0000 (15:32 +0200)]
Merge branch 'childless'

Adds support for childless initiation of IKE_SAs (RFC 6023) e.g. to
force a separate DH exchange for all CHILD_SAs including the first one.

Also allows the initiation of only the IKE_SA via swanctl --initiate if
the peer supports this extension.

Closes strongswan/strongswan#99.

3 years agotesting: Add swanctl/net2net-childless scenario
Tobias Brunner [Wed, 3 Apr 2019 09:15:50 +0000 (11:15 +0200)]
testing: Add swanctl/net2net-childless scenario

3 years agounit-tests: Add unit tests for childless IKE_SA initiation
Tobias Brunner [Tue, 2 Apr 2019 14:44:44 +0000 (16:44 +0200)]
unit-tests: Add unit tests for childless IKE_SA initiation

3 years agounit-tests: Make childless initiation configurable
Tobias Brunner [Tue, 2 Apr 2019 14:24:01 +0000 (16:24 +0200)]
unit-tests: Make childless initiation configurable

3 years agounit-tests: Add helper to create but not yet establish two IKE_SAs
Tobias Brunner [Tue, 2 Apr 2019 14:23:34 +0000 (16:23 +0200)]
unit-tests: Add helper to create but not yet establish two IKE_SAs

3 years agounit-tests: Add macros to assert certain payloads are (not) in a message
Tobias Brunner [Tue, 2 Apr 2019 14:22:21 +0000 (16:22 +0200)]
unit-tests: Add macros to assert certain payloads are (not) in a message

3 years agovici: Support initiation of IKE_SAs
Tobias Brunner [Fri, 29 Mar 2019 16:38:39 +0000 (17:38 +0100)]
vici: Support initiation of IKE_SAs

The configuration must allow the initiation of a childless IKE_SA (which
is already the case with the default of 'accept').

3 years agovici: Make childless initiation of IKE_SAs configurable
Tobias Brunner [Fri, 29 Mar 2019 16:13:49 +0000 (17:13 +0100)]
vici: Make childless initiation of IKE_SAs configurable

3 years agocontroller: Make child config optional for initiate()
Tobias Brunner [Fri, 29 Mar 2019 15:50:26 +0000 (16:50 +0100)]
controller: Make child config optional for initiate()

3 years agochild-create: Initiate and handle childless IKE_SAs according to RFC 6023
Tobias Brunner [Fri, 29 Mar 2019 15:46:59 +0000 (16:46 +0100)]
child-create: Initiate and handle childless IKE_SAs according to RFC 6023

3 years agoike-init: Notify initiator if childless IKE_SAs are accepted
Tobias Brunner [Fri, 29 Mar 2019 14:18:08 +0000 (15:18 +0100)]
ike-init: Notify initiator if childless IKE_SAs are accepted

3 years agoike-cfg: Add setting for childless IKE_SAs
Tobias Brunner [Fri, 29 Mar 2019 14:06:20 +0000 (15:06 +0100)]
ike-cfg: Add setting for childless IKE_SAs

3 years agoike-cfg: Pass arguments as struct
Tobias Brunner [Fri, 29 Mar 2019 11:11:10 +0000 (12:11 +0100)]
ike-cfg: Pass arguments as struct

3 years agoproposal-substructure: Fix incorrect type for IKEv2 proposals
SophieK [Thu, 25 Apr 2019 06:39:32 +0000 (14:39 +0800)]
proposal-substructure: Fix incorrect type for IKEv2 proposals

Luckily, the type is only used once when generating payloads and there it
doesn't matter because the encoding rules are the same.

Closes strongswan/strongswan#135.

3 years agotesting: Use renamed systemd unit
Tobias Brunner [Fri, 29 Mar 2019 09:24:07 +0000 (10:24 +0100)]
testing: Use renamed systemd unit

While the alias is available after enabling the unit, we don't
actually do that in our testing environment (adding a symlink manually
would work too, then again, why not just use the proper name?).

3 years agoinit: Rename systemd units
Tobias Brunner [Fri, 29 Mar 2019 09:00:42 +0000 (10:00 +0100)]
init: Rename systemd units

Use strongswan-starter for the legacy unit and simply strongswan for the
modern one (strongswan-swanctl is configured as alias, which should
cause the installation of symlinks when the service is enabled via

3 years agoMerge branch 'wolfssl'
Tobias Brunner [Wed, 24 Apr 2019 11:55:20 +0000 (13:55 +0200)]
Merge branch 'wolfssl'

Adds a plugin that uses wolfSSL for cryptographic operations.

Closes strongswan/strongswan#133.

3 years agotravis: Run tests against wolfSSL
Tobias Brunner [Tue, 9 Apr 2019 08:40:54 +0000 (10:40 +0200)]
travis: Run tests against wolfSSL

Check for wolfssl/options.h because if it isn't included, checking other
headers will trigger a warning about hardening the wolfSSL build, which
will cause the check to fail with -Werror.

If the file doesn't exist because user_settings.h is used, the check may
be skipped by configuring with `ac_cv_header_wolfssl_options_h=yes`.

3 years agowolfssl: Fixes, code style changes and some refactorings
Tobias Brunner [Fri, 5 Apr 2019 10:03:18 +0000 (12:03 +0200)]
wolfssl: Fixes, code style changes and some refactorings

The main fixes are

 * the generation of fingerprints for RSA, ECDSA, and EdDSA
 * the encoding of ECDSA private keys
 * calculating p and q for RSA private keys
 * deriving the public key for raw Ed25519 private keys

Also, instead of numeric literals for buffer lengths ASN.1 related
constants are used.

3 years agounit-tests: Add tests for ECDSA fingerprints and encoding
Tobias Brunner [Mon, 8 Apr 2019 15:57:19 +0000 (17:57 +0200)]
unit-tests: Add tests for ECDSA fingerprints and encoding

3 years agounit-tests: Add tests for RSA fingerprints and encoding
Tobias Brunner [Mon, 8 Apr 2019 10:23:36 +0000 (12:23 +0200)]
unit-tests: Add tests for RSA fingerprints and encoding

3 years agochunk: Add helper to copy a chunk left-padded to a certain length
Tobias Brunner [Mon, 8 Apr 2019 13:24:23 +0000 (15:24 +0200)]
chunk: Add helper to copy a chunk left-padded to a certain length

3 years agowolfssl: Add wolfSSL plugin for cryptographic implementations
Sean Parkinson [Wed, 3 Apr 2019 07:06:34 +0000 (17:06 +1000)]
wolfssl: Add wolfSSL plugin for cryptographic implementations

3 years agoMerge branch 'android-fixes'
Tobias Brunner [Wed, 24 Apr 2019 09:36:26 +0000 (11:36 +0200)]
Merge branch 'android-fixes'

Fixes an upgrade issue and includes UTF8 support for EAP-MSCHAPv2.

3 years agosocket-default: Fix setting DSCP value on FreeBSD
Tobias Brunner [Tue, 23 Apr 2019 09:49:04 +0000 (11:49 +0200)]
socket-default: Fix setting DSCP value on FreeBSD

Fixes #3030.

3 years agoandroid: New release after fixing DB update and adding UTF-8 for EAP-MSCHAPv2
Tobias Brunner [Tue, 16 Apr 2019 13:58:31 +0000 (15:58 +0200)]
android: New release after fixing DB update and adding UTF-8 for EAP-MSCHAPv2

3 years agoandroid: Fix database upgrade from older versions
Tobias Brunner [Mon, 15 Apr 2019 12:24:23 +0000 (14:24 +0200)]
android: Fix database upgrade from older versions

3 years agoeap-mschapv2: Convert UTF-8-encoded passwords
Tobias Brunner [Wed, 10 Apr 2019 15:38:35 +0000 (17:38 +0200)]
eap-mschapv2: Convert UTF-8-encoded passwords

Instead of assuming passwords are simply ASCII-encoded we now assume they are
provided UTF-8-encoded, which is quite likely nowadays.  The UTF-8 byte
sequences are not validated, however, only valid code points are encoded
as UTF-16LE.

Fixes #3014.

3 years agotesting: Use latest tkm-rpc and x509-ada versions
Tobias Brunner [Mon, 15 Apr 2019 16:31:12 +0000 (18:31 +0200)]
testing: Use latest tkm-rpc and x509-ada versions

Includes fixes for larger signatures, critical extensions and
utf8Strings in DNs.

3 years agochild-create: Make sure the mode selected by the responder is acceptable
Tobias Brunner [Tue, 9 Apr 2019 17:13:53 +0000 (19:13 +0200)]
child-create: Make sure the mode selected by the responder is acceptable

Previously, the initiator would install the SA in transport mode if the
peer sent back the USE_TRANSPORT_MODE notify, even if that was not
requested originally.

3 years agomessage: Enforce encryption except for INFORMATIONALs
Tobias Brunner [Tue, 9 Apr 2019 09:42:19 +0000 (11:42 +0200)]
message: Enforce encryption except for INFORMATIONALs

The only messages that are generally sent encrypted but could be sent
unencrypted are INFORMATIONALs (currently only used for IKEv1 and ME
connectivity checks).  This should prevent issues if the keymat_t behaves
incorrectly and does not return an aead_t when it actually should.

3 years agotesting: Create new files in mounted strongSwan sources as regular user
Tobias Brunner [Mon, 15 Apr 2019 11:23:05 +0000 (13:23 +0200)]
testing: Create new files in mounted strongSwan sources as regular user

3 years agoike-sa-manager: Extract IKE SPI labeling feature from charon-tkm
Tobias Brunner [Tue, 9 Apr 2019 16:29:10 +0000 (18:29 +0200)]
ike-sa-manager: Extract IKE SPI labeling feature from charon-tkm

Might be useful for users of other daemons too. Note that compared to the
previous implementation in charon-tkm, the mask/label are applied in
network order.

Closes strongswan/strongswan#134.

3 years agoeap-aka-3gpp: Ignore test runner in repository
Tobias Brunner [Thu, 4 Apr 2019 15:56:41 +0000 (17:56 +0200)]
eap-aka-3gpp: Ignore test runner in repository

3 years agotravis: Check for unignored build artifacts after the build
Tobias Brunner [Thu, 4 Apr 2019 15:41:53 +0000 (17:41 +0200)]
travis: Check for unignored build artifacts after the build

3 years agoike-config: If we don't send a CFG_REQUEST, we don't expect a CFG_REPLY
Tobias Brunner [Fri, 29 Mar 2019 10:05:42 +0000 (11:05 +0100)]
ike-config: If we don't send a CFG_REQUEST, we don't expect a CFG_REPLY

Previously, attributes in an incorrectly sent CFG_REPLY would still be passed
to attribute handlers.  This does not prevent handlers from receiving
unrequested attributes if they requested at least one other.

3 years agoike-config: Ignore unrequested virtual IP addresses
Tobias Brunner [Thu, 28 Mar 2019 17:44:08 +0000 (18:44 +0100)]
ike-config: Ignore unrequested virtual IP addresses

But forward them to handlers in case they requested them.

3 years agoMerge branch 'xfrmi'
Tobias Brunner [Thu, 4 Apr 2019 08:41:01 +0000 (10:41 +0200)]
Merge branch 'xfrmi'

This adds support for XFRM interfaces, which replace VTI devices and are
available with 4.19+ Linux kernels.

IPsec SAs and policies are associated with such interfaces via interface
IDs that can be configured on the CHILD_SA-level (dynamic IDs may
optionally be allocated for each instance and even direction) or on the
IKE_SA-level (again, dynamic IDs may be optionally allocated per IKE_SA).
IDs on an IKE_SA are inherited by all CHILD_SAs created under it, unless
the child configuration overrides them.

The effect the interface ID has on policies is similar to that of marks,
i.e. they won't match packets unless they are routed via interface with
matching interface ID.  So it's possible to negotiate e.g. as
traffic selector on both sides and then control the affected traffic via

It's possible to use separate interfaces for in- and outbound traffic (or
only use an interface in one direction and regular policies in the other).

Since iproute2 does not yet support XFRM interfaces, a small utility is
provided that allows creating and listing XFRM interfaces.

Interfaces may be created dynamically via updown/vici scripts or
statically (before or after establishing the SAs).  Routes must be added
manually as needed (the daemon will not install any routes for outbound
policies with an interface ID).

When moving XFRM interfaces to other network namespaces they retain access
to the SAs and policies created in the original namespace, which allows
providing IPsec tunnels for processes in other network namespaces without
giving them access to the IPsec keys or IKE credentials.

Fixes #2845.

3 years agotesting: Add scenario that uses IKE-specific interface IDs
Tobias Brunner [Mon, 25 Mar 2019 17:31:28 +0000 (18:31 +0100)]
testing: Add scenario that uses IKE-specific interface IDs

3 years agotesting: Install python-daemon with strongSwan for use in updown scripts
Tobias Brunner [Wed, 27 Mar 2019 10:13:41 +0000 (11:13 +0100)]
testing: Install python-daemon with strongSwan for use in updown scripts

3 years agotesting: Add /etc/resolv.conf when building strongSwan
Tobias Brunner [Wed, 27 Mar 2019 10:04:57 +0000 (11:04 +0100)]
testing: Add /etc/resolv.conf when building strongSwan

3 years agotesting: Enable Python eggs in testing environment (i.e. vici's Python bindings)
Tobias Brunner [Mon, 25 Mar 2019 13:28:21 +0000 (14:28 +0100)]
testing: Enable Python eggs in testing environment (i.e. vici's Python bindings)

3 years agoconfigure: Fix package version for python packages for developer releases
Tobias Brunner [Mon, 25 Mar 2019 13:26:11 +0000 (14:26 +0100)]
configure: Fix package version for python packages for developer releases

According to PEP 440 the suffix for development releases is .devN and
not just devN.

3 years agovici: Add support for interface ID configurable on IKE_SA
Tobias Brunner [Fri, 22 Mar 2019 17:33:46 +0000 (18:33 +0100)]
vici: Add support for interface ID configurable on IKE_SA

3 years agochild-sa: Pass default interface ID inherited from IKE_SA
Tobias Brunner [Fri, 22 Mar 2019 16:39:47 +0000 (17:39 +0100)]
child-sa: Pass default interface ID inherited from IKE_SA

Also pass optional arguments as struct.

3 years agoike-sa: Add property for interface ID
Tobias Brunner [Fri, 22 Mar 2019 16:19:51 +0000 (17:19 +0100)]
ike-sa: Add property for interface ID

3 years agoipsec-types: Move allocation of unique interface IDs to helper function
Tobias Brunner [Fri, 22 Mar 2019 16:18:00 +0000 (17:18 +0100)]
ipsec-types: Move allocation of unique interface IDs to helper function

3 years agopeer-cfg: Add property for interface ID
Tobias Brunner [Fri, 22 Mar 2019 15:23:59 +0000 (16:23 +0100)]
peer-cfg: Add property for interface ID

3 years agoxfrmi: Use helpers for nested Netlink attributes
Tobias Brunner [Fri, 8 Mar 2019 14:22:58 +0000 (15:22 +0100)]
xfrmi: Use helpers for nested Netlink attributes

3 years agokernel-netlink: Add helper functions for nested attributes
Tobias Brunner [Fri, 8 Mar 2019 14:20:40 +0000 (15:20 +0100)]
kernel-netlink: Add helper functions for nested attributes

3 years agotesting: Add scenarios that use XFRM interfaces
Tobias Brunner [Fri, 8 Feb 2019 17:57:38 +0000 (18:57 +0100)]
testing: Add scenarios that use XFRM interfaces

The network namespace scenario requires a kernel patch in 4.19 and 4.20
kernels (the fix is included in 5.0 kernels).

3 years agokernel-netlink: Don't install routes for CHILD_SAs with interface ID
Tobias Brunner [Fri, 1 Mar 2019 09:19:32 +0000 (10:19 +0100)]
kernel-netlink: Don't install routes for CHILD_SAs with interface ID

3 years agotesting: Enable XFRM interfaces and network namespaces in 4.19 and 5.0 kernel
Tobias Brunner [Tue, 19 Feb 2019 09:16:31 +0000 (10:16 +0100)]
testing: Enable XFRM interfaces and network namespaces in 4.19 and 5.0 kernel

3 years agoxfrmi: Move to a separate directory to fix monolithic build
Tobias Brunner [Wed, 13 Feb 2019 09:33:59 +0000 (10:33 +0100)]
xfrmi: Move to a separate directory to fix monolithic build

3 years agokernel-netlink: Add --list option to XFRM interfaces utility
Tobias Brunner [Tue, 12 Feb 2019 17:26:43 +0000 (18:26 +0100)]
kernel-netlink: Add --list option to XFRM interfaces utility

3 years agoswanctl: Report interface IDs in --list-sas
Tobias Brunner [Tue, 12 Feb 2019 15:20:12 +0000 (16:20 +0100)]
swanctl: Report interface IDs in --list-sas

3 years agovici: Report interface IDs
Tobias Brunner [Tue, 12 Feb 2019 11:13:41 +0000 (12:13 +0100)]
vici: Report interface IDs

3 years agoupdown: Pass interface ID to updown script
Tobias Brunner [Tue, 12 Feb 2019 11:11:10 +0000 (12:11 +0100)]
updown: Pass interface ID to updown script

3 years agoike: Reuse interface ID during CHILD_SA rekeyings
Tobias Brunner [Tue, 12 Feb 2019 10:59:38 +0000 (11:59 +0100)]
ike: Reuse interface ID during CHILD_SA rekeyings

3 years agochild-sa: Configure interface ID on SAs and policies
Tobias Brunner [Tue, 12 Feb 2019 10:39:11 +0000 (11:39 +0100)]
child-sa: Configure interface ID on SAs and policies

3 years agovici: Make interface ID configurable
Tobias Brunner [Tue, 12 Feb 2019 11:10:06 +0000 (12:10 +0100)]
vici: Make interface ID configurable

3 years agoipsec-types: Add helper to parse interface ID
Tobias Brunner [Tue, 12 Feb 2019 11:25:53 +0000 (12:25 +0100)]
ipsec-types: Add helper to parse interface ID

3 years agokernel-interface: Consider interface ID when allocating reqids
Tobias Brunner [Tue, 12 Feb 2019 10:10:04 +0000 (11:10 +0100)]
kernel-interface: Consider interface ID when allocating reqids

3 years agochild-cfg: Add property for interface ID
Tobias Brunner [Tue, 12 Feb 2019 09:19:36 +0000 (10:19 +0100)]
child-cfg: Add property for interface ID

3 years agokernel-netlink: Make interface ID configurable on SAs and policies
Tobias Brunner [Fri, 1 Feb 2019 16:29:20 +0000 (17:29 +0100)]
kernel-netlink: Make interface ID configurable on SAs and policies

3 years agoinclude: Add XFRM attribute identifier for interface ID
Tobias Brunner [Fri, 1 Feb 2019 12:53:29 +0000 (13:53 +0100)]
include: Add XFRM attribute identifier for interface ID

3 years agokernel-netlink: Add utility to create XFRM interfaces
Tobias Brunner [Fri, 1 Feb 2019 08:28:10 +0000 (09:28 +0100)]
kernel-netlink: Add utility to create XFRM interfaces

This is mainly to see what's necessary to create them (in case we
integrate this into the daemon) and to experiment in our testing
environment without having to add a patched version of iproute2 (the
4.20.0 version in stretch-backports doesn't support XFRM interfaces
yet).  The regular version of iproute2 can be used for other operations
with these interfaces (delete, up, addrs etc.).

3 years agokernel-netlink: Calculate length of Netlink messages correctly
Tobias Brunner [Fri, 1 Feb 2019 08:19:03 +0000 (09:19 +0100)]
kernel-netlink: Calculate length of Netlink messages correctly

This is relevant when adding e.g. strings whose length are not aligned.

3 years agoUse Botan 2.10.0 for tests
Tobias Brunner [Mon, 1 Apr 2019 09:01:46 +0000 (11:01 +0200)]
Use Botan 2.10.0 for tests

3 years agoVersion bump to 5.8.0dr2 5.8.0dr2
Andreas Steffen [Sat, 30 Mar 2019 16:11:34 +0000 (17:11 +0100)]
Version bump to 5.8.0dr2

3 years agotesting: Updated expired certificates
Andreas Steffen [Sat, 30 Mar 2019 16:10:51 +0000 (17:10 +0100)]
testing: Updated expired certificates

3 years agoTesting: Removed tnc/tnccs-20-server-retry scenario
Andreas Steffen [Fri, 29 Mar 2019 15:54:54 +0000 (16:54 +0100)]
Testing: Removed tnc/tnccs-20-server-retry scenario

3 years agoCorrected use of PB-TNC CRETRY and SRETRY batches
Andreas Steffen [Fri, 29 Mar 2019 14:33:24 +0000 (15:33 +0100)]
Corrected use of PB-TNC CRETRY and SRETRY batches

The PB-TNC finite state machine according to section 3.2 of RFC 5793
was not correctly implemented when sending either a CRETRY or SRETRY
batch. These batches can only be sent in the "Decided" state and a
CRETRY batch can immediately carry all messages usually transported
by a CDATA batch. strongSwan currently is not able to send a SRETRY
batch since full-duplex mode for PT-TLS isn't supported yet.

3 years agotesting: Disable gcrypt plugin for swanctl
Tobias Brunner [Thu, 28 Mar 2019 17:16:56 +0000 (18:16 +0100)]
testing: Disable gcrypt plugin for swanctl

Sometimes swanctl hangs when initializing the plugin and it apparently
gathers entropy.