]> git.ipfire.org Git - thirdparty/strongswan.git/log
thirdparty/strongswan.git
3 years agoswanctl: Clear config from memory
Tobias Brunner [Thu, 30 Sep 2021 13:16:48 +0000 (15:16 +0200)] 
swanctl: Clear config from memory

The 'secrets' section in the config file may contain passwords/PINs.

3 years agosettings: Add destructor that wipes contents
Tobias Brunner [Thu, 30 Sep 2021 13:13:35 +0000 (15:13 +0200)] 
settings: Add destructor that wipes contents

3 years agostroke: Clear ipsec.secrets file from memory
Tobias Brunner [Thu, 30 Sep 2021 13:38:38 +0000 (15:38 +0200)] 
stroke: Clear ipsec.secrets file from memory

3 years agoswanctl: Clear possibly unencrypted key/container files from memory
Tobias Brunner [Thu, 30 Sep 2021 13:37:46 +0000 (15:37 +0200)] 
swanctl: Clear possibly unencrypted key/container files from memory

3 years agopem: Clear possibly unencrypted key file from memory
Tobias Brunner [Thu, 30 Sep 2021 13:33:40 +0000 (15:33 +0200)] 
pem: Clear possibly unencrypted key file from memory

3 years agochunk: Optionally clear mmap'd chunk before unmapping
Tobias Brunner [Thu, 30 Sep 2021 10:39:08 +0000 (12:39 +0200)] 
chunk: Optionally clear mmap'd chunk before unmapping

This is mostly for the non-mmap case as with mmap available, access to the
unmapped memory isn't easily possible (e.g. opening the same area with
MAP_ANONYMOUS | MAP_UNINITIALIZED is usually prevented by the missing
CONFIG_MMAP_ALLOW_UNINITIALIZED option in most kernels).

3 years agovici: Clear all request messages in case they contain secrets
Tobias Brunner [Thu, 30 Sep 2021 10:22:59 +0000 (12:22 +0200)] 
vici: Clear all request messages in case they contain secrets

3 years agovici: Clear cached strings in case the message contained shared secrets
Tobias Brunner [Thu, 30 Sep 2021 10:10:44 +0000 (12:10 +0200)] 
vici: Clear cached strings in case the message contained shared secrets

3 years agopem: Clear chunks after decrypting files
Tobias Brunner [Thu, 30 Sep 2021 10:03:42 +0000 (12:03 +0200)] 
pem: Clear chunks after decrypting files

3 years agoshared-key: Clear shared secret when destroyed
Tobias Brunner [Thu, 30 Sep 2021 10:00:25 +0000 (12:00 +0200)] 
shared-key: Clear shared secret when destroyed

3 years agoMerge branch 'testing-bullseye'
Tobias Brunner [Fri, 1 Oct 2021 15:13:18 +0000 (17:13 +0200)] 
Merge branch 'testing-bullseye'

Use Debian bullseye as base image for the testing environment.

3 years agotesting: Use Debian bullseye as base image
Tobias Brunner [Tue, 21 Sep 2021 17:30:00 +0000 (19:30 +0200)] 
testing: Use Debian bullseye as base image

3 years agotesting: Determine Debian version in TNC scenarios automatically again
Tobias Brunner [Wed, 22 Sep 2021 12:10:54 +0000 (14:10 +0200)] 
testing: Determine Debian version in TNC scenarios automatically again

Was statically set to 10 since f3d96b7bc9e2 ("Version bump to 5.9.1dr1").

3 years agolibimcv: Add Debian 11 (bullseye) to IMV database
Tobias Brunner [Wed, 22 Sep 2021 12:04:39 +0000 (14:04 +0200)] 
libimcv: Add Debian 11 (bullseye) to IMV database

3 years agotesting: Replace `arp` with `ip neigh`
Tobias Brunner [Wed, 22 Sep 2021 11:46:19 +0000 (13:46 +0200)] 
testing: Replace `arp` with `ip neigh`

The net-tools package, which provides arp and other older utilities like
ifconfig, is not installed anymore on Debian bullseye.

3 years agotesting: Allow DNS via TCP in net2net-dnscert scenario
Tobias Brunner [Wed, 22 Sep 2021 11:28:53 +0000 (13:28 +0200)] 
testing: Allow DNS via TCP in net2net-dnscert scenario

New versions of Bind limit the maximum UDP message size to 1232 bytes,
which is the same that newer versions of libunbound propose as maximum via
EDNS in requests, so increasing the limit on the server wouldn't help.

Instead we allow DNS via TCP so the client can switch after receiving the
truncated UDP response.

3 years agotesting: Remove support for Debian jessie and FreeRADIUS 2
Tobias Brunner [Tue, 21 Sep 2021 09:18:16 +0000 (11:18 +0200)] 
testing: Remove support for Debian jessie and FreeRADIUS 2

3 years agotesting: Support build with Debian bullseye base image
Tobias Brunner [Tue, 21 Sep 2021 09:15:43 +0000 (11:15 +0200)] 
testing: Support build with Debian bullseye base image

3 years agoopenssl: Use correct key types when adopting Ed25519/448 keys
Tobias Brunner [Mon, 20 Sep 2021 15:20:30 +0000 (17:20 +0200)] 
openssl: Use correct key types when adopting Ed25519/448 keys

Fixes: cbf07ab5b0b6 ("openssl: Add support for Ed25519/Ed448")
3 years agoopenssl: Remove workaround for Brainpool ECDH curves for older OpenSSL versions
Tobias Brunner [Mon, 20 Sep 2021 14:31:33 +0000 (16:31 +0200)] 
openssl: Remove workaround for Brainpool ECDH curves for older OpenSSL versions

Using the workaround with the EVP interface, which we use to derive shared
keys since 74e02ff5e624 ("openssl: Mainly use EVP interface for ECDH"),
would actually require us to register the OIDs of these curves as NID.
Otherwise, the two EC_GROUPs used by private and public key objects
are not considered the same and the key derivation fails.

Since the curves are supported by OpenSSL since 1.0.2 it's probably rare to
find a version without them nowadays.  One exception is the old BoringSSL
version we still use on Android, which defines the NIDs but not the curve
data.  However, that version also lacks support to register OIDs as NIDs,
so the only option to support these groups there would be to got back to not
using the EVP interface, which isn't in anyone's interest.  If there really
is a need for them there, we could probably patch BoringSSL or use OpenSSL.

3 years agogithub: Fix build on macOS due to moved OpenSSL location
Tobias Brunner [Thu, 30 Sep 2021 14:21:45 +0000 (16:21 +0200)] 
github: Fix build on macOS due to moved OpenSSL location

3 years agocrl: Fix compilation with OpenSSL 3.0
Tobias Brunner [Thu, 30 Sep 2021 07:40:47 +0000 (09:40 +0200)] 
crl: Fix compilation with OpenSSL 3.0

Closes strongswan/strongswan#635

3 years agogithub: Use OpenSSL 1.1 on macOS
Tobias Brunner [Thu, 23 Sep 2021 14:56:19 +0000 (16:56 +0200)] 
github: Use OpenSSL 1.1 on macOS

The "openssl" alias now defaults to OpenSSL 3.0, which produces a lot of
deprecation warnings.  To avoid build failures due to `-Werror`, stay with
OpenSSL 1.1 until we can get rid of these issues.

3 years agokernel-pfkey: Wipe request/response messages when managing SAs
Tobias Brunner [Thu, 23 Sep 2021 13:24:52 +0000 (15:24 +0200)] 
kernel-pfkey: Wipe request/response messages when managing SAs

3 years agoVersion bump to 5.9.4dr2 5.9.4dr2
Andreas Steffen [Tue, 21 Sep 2021 12:18:16 +0000 (14:18 +0200)] 
Version bump to 5.9.4dr2

3 years agoChange shebang for Python scripts to use python3
Tobias Brunner [Tue, 21 Sep 2021 09:13:50 +0000 (11:13 +0200)] 
Change shebang for Python scripts to use python3

3 years agotesting: Fix updown script in route-based/net2net-xfrmi-ike scenario
Tobias Brunner [Tue, 21 Sep 2021 10:40:52 +0000 (12:40 +0200)] 
testing: Fix updown script in route-based/net2net-xfrmi-ike scenario

With the update to Python 3 the encoding of the values in vici messages
changed to bytestrings (the keys are properly decoded).  And getting the
first CHILD_SA also needs a change.

The logger is now also initialized after daemonizing to avoid that opened
sockets are closed etc.

3 years agotesting: Switch to Python 3
Andreas Steffen [Tue, 21 Sep 2021 07:28:49 +0000 (09:28 +0200)] 
testing: Switch to Python 3

Both swidGenerator and strongTNC were migrated to Python 3.
This allows to migrate all testing scripts to Python 3, too.

3 years agokeymat_v2: Properly wipe DH secret during IKE_SA rekeying
Tobias Brunner [Mon, 20 Sep 2021 14:08:00 +0000 (16:08 +0200)] 
keymat_v2: Properly wipe DH secret during IKE_SA rekeying

While `secret` is wiped explicitly, it wasn't when concatenating with
the nonces.

3 years agosec-updater: Use latest versions
Andreas Steffen [Tue, 7 Sep 2021 16:55:02 +0000 (18:55 +0200)] 
sec-updater: Use latest versions

3 years agoos_info: Parse /etc/os-release first
Andreas Steffen [Tue, 7 Sep 2021 12:59:36 +0000 (14:59 +0200)] 
os_info: Parse /etc/os-release first

3 years agogithub: Remove build on Ubuntu Xenial (16.04)
Tobias Brunner [Wed, 8 Sep 2021 10:00:12 +0000 (12:00 +0200)] 
github: Remove build on Ubuntu Xenial (16.04)

Support for it will be removed later this month.

3 years agovici: Update supported Python versions
Tobias Brunner [Wed, 8 Sep 2021 09:55:41 +0000 (11:55 +0200)] 
vici: Update supported Python versions

3 years agovici: Add DBG4 messages that print loaded shared keys and PINs
Noel Kuntze [Fri, 27 Aug 2021 17:05:03 +0000 (19:05 +0200)] 
vici: Add DBG4 messages that print loaded shared keys and PINs

Closes strongswan/strongswan#217

3 years agoike: Initiate new IKE_SA not until all children are queued
Tobias Brunner [Mon, 12 Jul 2021 12:47:57 +0000 (14:47 +0200)] 
ike: Initiate new IKE_SA not until all children are queued

If there are many CHILD_SAs, the time between initiating the new IKE_SA
and checking it in might be longer (depending on what else is going on
in the daemon) than the retransmission timeout and no retransmits might
be sent afterwards for this SA (it will just linger around dead).
Calling initiate() last should avoid that (we do this similarly for MBB
reauthentication).

3 years agoikev2: Only request reauth during IKE_AUTH if active reauth is not possible
Tobias Brunner [Tue, 29 Jun 2021 13:54:15 +0000 (15:54 +0200)] 
ikev2: Only request reauth during IKE_AUTH if active reauth is not possible

If we can initiate the reauthentication ourselves, there is no reason to
explicitly request the peer to do so (at basically the same time).

3 years agoike-rekey: Respond with TEMPORARY_FAILURE while reauthenticating
Tobias Brunner [Tue, 29 Jun 2021 13:50:43 +0000 (15:50 +0200)] 
ike-rekey: Respond with TEMPORARY_FAILURE while reauthenticating

This could lead to duplicates as the new IKE_SA can't delete the old one
once reauthentication is complete if it was replaced by a rekeying.

3 years agoike: Don't rekey IKE_SA while reauthenticating
Tobias Brunner [Tue, 29 Jun 2021 13:25:48 +0000 (15:25 +0200)] 
ike: Don't rekey IKE_SA while reauthenticating

If we are using make-before-break reauthentication, this could lead to
duplicates as the new IKE_SA wouldn't be able to delete the previous
one if it was replaced by a rekeying.

3 years agoike-delete: Don't call reestablish() when reauthenticating
Tobias Brunner [Wed, 16 Jun 2021 11:54:18 +0000 (13:54 +0200)] 
ike-delete: Don't call reestablish() when reauthenticating

If we initiated a make-before-break reauthentication and the peer
concurrently deletes the IKE_SA (e.g. because it uses break-before-make
reauthentication), we would create a duplicate IKE_SA (the condition forces
a recreation of all existing CHILD_SAs because reestablish() is also called
to complete a break-before-make reauthentication).

3 years agoopenssl: Improve compatibility to newer BoringSSL releases
Tobias Brunner [Tue, 22 Jun 2021 08:55:31 +0000 (10:55 +0200)] 
openssl: Improve compatibility to newer BoringSSL releases

Closes strongswan/strongswan#352

3 years agoMerge branch 'swanctl-ssh-public-keys'
Tobias Brunner [Mon, 23 Aug 2021 16:01:49 +0000 (18:01 +0200)] 
Merge branch 'swanctl-ssh-public-keys'

Minor changes that allow loading SSH public keys via `pubkeys` in
swanctl.conf.

References strongswan/strongswan#467

3 years agopubkey: Ignore X.509 flags passed to the parser
Tobias Brunner [Tue, 22 Jun 2021 08:35:10 +0000 (10:35 +0200)] 
pubkey: Ignore X.509 flags passed to the parser

3 years agopubkey: Don't assume blobs are only ASN.1 DER
Tobias Brunner [Mon, 21 Jun 2021 09:44:51 +0000 (11:44 +0200)] 
pubkey: Don't assume blobs are only ASN.1 DER

Also forward the blob's type when parsing pubkeys.

3 years agovici: Use the more generic BUILD_BLOB to parse certificates/public keys
Tobias Brunner [Mon, 21 Jun 2021 09:39:07 +0000 (11:39 +0200)] 
vici: Use the more generic BUILD_BLOB to parse certificates/public keys

3 years agogithub: Remove github.ref from cache keys
Tobias Brunner [Fri, 20 Aug 2021 09:42:39 +0000 (11:42 +0200)] 
github: Remove github.ref from cache keys

According to the documentation for actions/cache, the lookup is already
scoped to the current branch (with fallback to any base branch including
the default branch).

3 years agotpm_extendpcr: Support platforms not having syslog()
Tobias Brunner [Mon, 23 Aug 2021 09:30:15 +0000 (11:30 +0200)] 
tpm_extendpcr: Support platforms not having syslog()

3 years agotpm_extendpcr: Build does not depend on tpm plugin but libtpmtss
Tobias Brunner [Mon, 23 Aug 2021 08:44:41 +0000 (10:44 +0200)] 
tpm_extendpcr: Build does not depend on tpm plugin but libtpmtss

The utility basically requires a TSS, the tpm plugin is not used at all.

3 years agoVersion bump to 5.9.4dr1 5.9.4dr1
Andreas Steffen [Sun, 22 Aug 2021 08:55:37 +0000 (10:55 +0200)] 
Version bump to 5.9.4dr1

3 years agolibtpmtss: Initialize library from all users
Tobias Brunner [Fri, 20 Aug 2021 14:34:48 +0000 (16:34 +0200)] 
libtpmtss: Initialize library from all users

Previously, only the tpm plugin initialized the library, so in order
to use a TPM 2.0 (a required TCTI library is loaded via init), it was
necessary to load it even if none of its actual features were used.

3 years agoconfigure: Enforce tss-tss2 if tpm plugin is enabled
Tobias Brunner [Fri, 20 Aug 2021 08:45:44 +0000 (10:45 +0200)] 
configure: Enforce tss-tss2 if tpm plugin is enabled

The plugin is useless without a TSS 2.0 and this is currently the only one.

3 years agoopenssl: Use a longer key to test/initialize HMAC instances
Tobias Brunner [Mon, 16 Aug 2021 08:50:01 +0000 (10:50 +0200)] 
openssl: Use a longer key to test/initialize HMAC instances

OpenSSL enforces a minimum of 14 bytes (112 bits) on the key size when
used in FIPS-mode (as required by SP 800-131A).  So by using an empty
string, instantiation always failed.  32 bytes (256 bits) should be safe
for now.

Closes strongswan/strongswan#557

3 years agotesting: Fix build of tkm and tkm-rpc
Tobias Brunner [Tue, 17 Aug 2021 17:20:45 +0000 (19:20 +0200)] 
testing: Fix build of tkm and tkm-rpc

3 years agogithub: Use latest version of SonarScanner
Tobias Brunner [Mon, 16 Aug 2021 09:24:14 +0000 (11:24 +0200)] 
github: Use latest version of SonarScanner

3 years agokernel-netlink: Initialize ifreq structs when detecting offload capability
Tobias Brunner [Wed, 11 Aug 2021 13:35:52 +0000 (15:35 +0200)] 
kernel-netlink: Initialize ifreq structs when detecting offload capability

4 years agogitignore: Add nbproject/
Noel Kuntze [Fri, 23 Jul 2021 21:29:23 +0000 (23:29 +0200)] 
gitignore: Add nbproject/

As used by the NetBeans IDE.

Closes strongswan/strongswan#505.

4 years agoIgnore android-* tags when using `git describe` android-2.3.3-1
Tobias Brunner [Wed, 14 Jul 2021 08:27:11 +0000 (10:27 +0200)] 
Ignore android-* tags when using `git describe`

4 years agoMerge branch 'android-updates' android-2.3.3
Tobias Brunner [Wed, 14 Jul 2021 08:01:45 +0000 (10:01 +0200)] 
Merge branch 'android-updates'

Adds a button to install user certificates and updates the target SDK
version for Android 11 (and the related deprecation fixes), which will
be mandatory later this year.

The release also includes an older commit that changed how DNS servers
are applied to TUN devices (cd10ae2ff050 ("android: Explicitly apply DNS
servers to the TUN device")).

Also added are metadata for F-Droid.

4 years agoandroid: New release after adding button to install PKCS#12 and SDK update
Tobias Brunner [Tue, 13 Jul 2021 13:32:05 +0000 (15:32 +0200)] 
android: New release after adding button to install PKCS#12 and SDK update

It also includes an unreleased change that affects when/how DNS servers
are applied to the TUN device.

4 years agoandroid: Request QUERY_ALL_PACKAGES permission
Tobias Brunner [Tue, 13 Jul 2021 13:05:29 +0000 (15:05 +0200)] 
android: Request QUERY_ALL_PACKAGES permission

This is required when targeting Android 11 (API 30) in order to see all
packages, which we use to allow selecting apps ex-/included from VPN
profiles and for the EAP-TNC use case.

4 years agoandroid: Add Triple-T metadata for F-Droid
Tobias Brunner [Mon, 7 Jun 2021 13:12:19 +0000 (15:12 +0200)] 
android: Add Triple-T metadata for F-Droid

We currently don't use Triple-T but the metadata is also parsed by F-Droid
to provide information for included apps.

4 years agoandroid: Add a button to install user certificates
Tobias Brunner [Mon, 12 Jul 2021 16:58:58 +0000 (18:58 +0200)] 
android: Add a button to install user certificates

Newer Android versions don't provide this option anymore on the
selection dialog.

4 years agoandroid: Replace deprecated Switch with SwitchCompat
Tobias Brunner [Tue, 13 Jul 2021 13:11:07 +0000 (15:11 +0200)] 
android: Replace deprecated Switch with SwitchCompat

4 years agoandroid: Replace deprecated AsyncTask instances
Tobias Brunner [Mon, 12 Jul 2021 15:58:16 +0000 (17:58 +0200)] 
android: Replace deprecated AsyncTask instances

As suggested by the Android docs, we use a global thread pool and handler
to avoid recreating them repeatedly.  Four threads should be more than
enough as we only use this to load CA certificates when the app starts
initially and to load user certs when editing a profile.

4 years agoandroid: Migrate from deprecated ViewPager to ViewPager2 for CA cert lists
Tobias Brunner [Mon, 12 Jul 2021 15:17:00 +0000 (17:17 +0200)] 
android: Migrate from deprecated ViewPager to ViewPager2 for CA cert lists

4 years agoandroid: Replace deprecated FileObserver() constructor with newer SDKs
Tobias Brunner [Mon, 12 Jul 2021 13:53:42 +0000 (15:53 +0200)] 
android: Replace deprecated FileObserver() constructor with newer SDKs

4 years agoandroid: Replace usage of deprecated Handler() constructor
Tobias Brunner [Mon, 12 Jul 2021 13:38:22 +0000 (15:38 +0200)] 
android: Replace usage of deprecated Handler() constructor

4 years agoandroid: Replace deprecated onActivityCreated() with onViewCreated()
Tobias Brunner [Fri, 9 Jul 2021 16:23:05 +0000 (18:23 +0200)] 
android: Replace deprecated onActivityCreated() with onViewCreated()

4 years agoandroid: Replace deprecated startActivityForResult/onActivityResult usage
Tobias Brunner [Fri, 9 Jul 2021 16:18:13 +0000 (18:18 +0200)] 
android: Replace deprecated startActivityForResult/onActivityResult usage

4 years agoandroid: Set compile-/targetSdkVersion to 30
Tobias Brunner [Mon, 28 Jun 2021 12:26:52 +0000 (14:26 +0200)] 
android: Set compile-/targetSdkVersion to 30

This will be mandatory for new apps in August and for existing apps
in November.  However, several classes like AsyncTask are now deprecated
so this needs some work to avoid warnings and problems in the future.

4 years agoandroid: Replace jcenter with mavenCentral repository
Tobias Brunner [Mon, 28 Jun 2021 12:26:24 +0000 (14:26 +0200)] 
android: Replace jcenter with mavenCentral repository

4 years agoandroid: Update dependencies
Tobias Brunner [Mon, 28 Jun 2021 12:22:07 +0000 (14:22 +0200)] 
android: Update dependencies

4 years agoandroid: Update Gradle plugin
Tobias Brunner [Mon, 28 Jun 2021 12:21:41 +0000 (14:21 +0200)] 
android: Update Gradle plugin

4 years agoUse wolfSSL 4.8.0 for tests
Tobias Brunner [Tue, 13 Jul 2021 08:19:56 +0000 (10:19 +0200)] 
Use wolfSSL 4.8.0 for tests

4 years agoVersion bump to 5.9.3 5.9.3
Andreas Steffen [Tue, 6 Jul 2021 12:00:39 +0000 (14:00 +0200)] 
Version bump to 5.9.3

4 years agovici: Suppress trailing nul character
Andreas Steffen [Tue, 6 Jul 2021 10:05:09 +0000 (12:05 +0200)] 
vici: Suppress trailing nul character

4 years agotesting: Use specific versions of swidGenerator and strongTNC
Tobias Brunner [Wed, 30 Jun 2021 13:16:47 +0000 (15:16 +0200)] 
testing: Use specific versions of swidGenerator and strongTNC

This way we get updated versions automatically (referencing "master"
required manually deleting the downloaded archives and the unpacked
directories).  It also allows switching versions when working in different
branches (note that REV can also be set to a commit ID, e.g. to test
changes before tagging them later and merging the branch).

4 years agokernel-netlink: Fix theoretical memory leak when parsing routes
Tobias Brunner [Fri, 25 Jun 2021 11:51:44 +0000 (13:51 +0200)] 
kernel-netlink: Fix theoretical memory leak when parsing routes

This currently can't happen as the kernel always puts RTA_TABLE as first
attribute in RTM_NEWROUTE messages.

4 years agoFixed some typos, courtesy of codespell
Tobias Brunner [Fri, 25 Jun 2021 09:32:29 +0000 (11:32 +0200)] 
Fixed some typos, courtesy of codespell

Main change is the conversion from the British cancelling/-ed to the
American canceling/-ed.

4 years agoVersion bump to 5.9.3rc1 5.9.3rc1
Andreas Steffen [Thu, 24 Jun 2021 07:18:54 +0000 (09:18 +0200)] 
Version bump to 5.9.3rc1

4 years agotesting: Build wolfSSL from the Git repository
Tobias Brunner [Tue, 22 Jun 2021 14:49:40 +0000 (16:49 +0200)] 
testing: Build wolfSSL from the Git repository

Use the same configure options etc. for both builds (no need for the cert
options as we don't use TLS or X.509 parsing) and switch to a Git commit
that includes the SHA-3 OID fix (it's actually the fix itself).

4 years agotesting: Fixed ikev2/farp scenario
Andreas Steffen [Tue, 22 Jun 2021 10:32:35 +0000 (12:32 +0200)] 
testing: Fixed ikev2/farp scenario

4 years agoVersion bump to 5.9.3dr4 5.9.3dr4
Andreas Steffen [Tue, 22 Jun 2021 08:33:07 +0000 (10:33 +0200)] 
Version bump to 5.9.3dr4

4 years agotesting: Migrate ikev2-stroke-bye scenarios to vici
Andreas Steffen [Wed, 26 May 2021 07:40:42 +0000 (09:40 +0200)] 
testing: Migrate ikev2-stroke-bye scenarios to vici

4 years agotesting: Fixed pretest script of ikev1/rw-psk-aggressive scenario
Andreas Steffen [Fri, 21 May 2021 10:04:43 +0000 (12:04 +0200)] 
testing: Fixed pretest script of ikev1/rw-psk-aggressive scenario

4 years agotesting: Migrate ikev2/host2host-transport-nat scenario to vici
Tobias Brunner [Thu, 17 Jun 2021 15:03:39 +0000 (17:03 +0200)] 
testing: Migrate ikev2/host2host-transport-nat scenario to vici

This also restores the test as it was before the referenced commit so it
again, as written in the description, demonstrates that venus is unable
to ping sun without IPsec tunnel.

Fixes: f27fb58ae0ec ("testing: Update description and test evaluation of host2host-transport-nat")
4 years agotesting: Migrate MOBIKE tests to vici
Tobias Brunner [Thu, 10 Jun 2021 14:41:01 +0000 (16:41 +0200)] 
testing: Migrate MOBIKE tests to vici

Note that the mobike-nat test has been removed as it basically did the same
as the mobike-virtual-ip-nat test.  Instead, the mobike-nat-mapping scenario
is added, which simulates a NAT router restart.

4 years agoike-mobike: Force MOBIKE update after NAT mappings changed
Tobias Brunner [Thu, 10 Jun 2021 14:39:18 +0000 (16:39 +0200)] 
ike-mobike: Force MOBIKE update after NAT mappings changed

The addresses observed by the client behind the NAT are exactly the same if
the NAT router gets restarted.

Fixes: 2b255f01afbc ("ike-mobike: Use ike_sa_t::update_hosts() to trigger events")
4 years agoike-sa: Log IKE endpoint changes
Tobias Brunner [Thu, 10 Jun 2021 14:38:23 +0000 (16:38 +0200)] 
ike-sa: Log IKE endpoint changes

4 years agoha: Register the correct IKE_SA with the manager after a rekeying
Tobias Brunner [Fri, 18 Jun 2021 13:31:06 +0000 (15:31 +0200)] 
ha: Register the correct IKE_SA with the manager after a rekeying

Fixes: 20dfbcad0875 ("ha: Register new IKE_SAs before calling inherit_post()")
Closes strongswan/strongswan#456.

4 years agoMerge branch 'vici-stuck'
Tobias Brunner [Mon, 21 Jun 2021 07:59:28 +0000 (09:59 +0200)] 
Merge branch 'vici-stuck'

Closes strongswan/strongswan#268.

4 years agovici: Signal waiting threads when skipping disconnected connections
Tobias Brunner [Fri, 7 May 2021 13:32:49 +0000 (15:32 +0200)] 
vici: Signal waiting threads when skipping disconnected connections

If two threads are waiting in find_entry() and remove_entry(),
respectively, and the former is woken first, the latter remains stuck
as it won't get signaled.

4 years agovici: Signal waiting threads when removing a connection entry
Tobias Brunner [Fri, 7 May 2021 08:53:57 +0000 (10:53 +0200)] 
vici: Signal waiting threads when removing a connection entry

If there are threads waiting in find_entry() and one in remove_entry()
and the latter is woken first by a thread calling put_entry(), the
former threads would remain stuck as they get never signaled.

4 years agokernel-netlink: Don't wait for VIPs to disappear during de-initialization
Tobias Brunner [Thu, 10 Jun 2021 09:34:37 +0000 (11:34 +0200)] 
kernel-netlink: Don't wait for VIPs to disappear during de-initialization

This can happen if an IKE_SA is terminated forcefully shortly before
terminating the daemon.  The thread that handles the terminate command
will call checkin_and_destroy(), which unregisters the IKE_SA from the
manager before destroying it.  The main thread that calls flush() on the
IKE_SA manager won't wait for this SA (its entry is already gone), so
the processor and in turn the watcher job/thread might get canceled
before the first thread started deleting the VIP.  It would then wait
indefinitely for a signal that can never be sent.

There is still a small chance the thread hangs in wait() if the state check
happens right before the watcher is canceled and it wasn't yet able to
deliver the event from the kernel, we counter that by rechecking the state
after a while.

4 years agoNEWS: Add news for 5.9.3
Tobias Brunner [Fri, 18 Jun 2021 08:31:31 +0000 (10:31 +0200)] 
NEWS: Add news for 5.9.3

4 years agotesting: Update Anet to version 0.4.2
Adrian-Ken Rueegsegger [Wed, 16 Jun 2021 16:27:23 +0000 (18:27 +0200)] 
testing: Update Anet to version 0.4.2

4 years agocharon-tkm: Remove useless checks when deriving IKE keys
Stefan Berghofer [Wed, 4 Mar 2020 16:58:21 +0000 (17:58 +0100)] 
charon-tkm: Remove useless checks when deriving IKE keys

4 years agocharon-tkm: Delegate encryption/decryption of IKE traffic to TKM
Stefan Berghofer [Tue, 23 Jul 2019 10:42:26 +0000 (12:42 +0200)] 
charon-tkm: Delegate encryption/decryption of IKE traffic to TKM

Co-authored-by: Tobias Brunner <tobias@strongswan.org>
4 years agopkcs11: Change how unavailable attributes like CKA_TRUSTED are handled
Tobias Brunner [Thu, 20 May 2021 14:52:49 +0000 (16:52 +0200)] 
pkcs11: Change how unavailable attributes like CKA_TRUSTED are handled

If a PKCS#11 library/token doesn't provide one or more attributes via
C_GetAttributeValue(), we get back CKR_ATTRIBUTE_TYPE_INVALID (similar
for protected attributes where CKR_ATTRIBUTE_SENSITIVE is returned).
This is not an error as the spec demands that all attributes have been
processed with the unavailable attributes having set their length
field to CK_UNAVAILABLE_INFORMATION.

We use this to handle the CKA_TRUSTED attribute, which some tokens
apparently don't support.  We previously used a version check to remove
the attribute from the call but even the latest spec doesn't make the
attribute mandatory (it's just in a list of "common" attributes for
CKO_CERTIFICATE objects, without a default value), so there are current
tokens that don't support it and prevent us from enumerating certificates.

4 years agoreceiver: Avoid division by 0 after system start if CLOCK_MONOTONIC is used
Tobias Brunner [Tue, 8 Jun 2021 13:33:36 +0000 (15:33 +0200)] 
receiver: Avoid division by 0 after system start if CLOCK_MONOTONIC is used

Depending on how CLOCK_MONOTONIC is implemented, time_monotonic() might
return 0 within 1 second after the system is started.  If that's the
case, we just default to 0 for now to avoid a crash (doesn't "hide" the
system time, but it's only the uptime anyway in this case).

Closes strongswan/strongswan#435.

4 years agocharon-nm: Simplify certificate enumeration and allow IDs other than DNs
Tobias Brunner [Wed, 9 Jun 2021 14:36:57 +0000 (16:36 +0200)] 
charon-nm: Simplify certificate enumeration and allow IDs other than DNs

This allows using SANs as identity instead of having to use the subject DN.

References strongswan/strongswan#437.