Using the source address to determine the interface is not correct for
net-to-net shunts between two interfaces on which the host has IP addresses
for each subnet.
Other threads are free to add/update/delete other policies.
This tries to prevent race conditions caused by releasing the mutex while
sending messages to the kernel. For instance, if break-before-make
reauthentication is used and one thread on the responder is delayed in
deleting the policies that another thread is concurrently adding for the
new SA. This could have resulted in no policies being installed
eventually.
Fixes#1400.
memeq() is currently used to compare these but if there is padding that
is not initialized the same for two instances the comparison fails.
Using this function ensures the objects are compared correctly.
If a pseudonym changed a new entry was added to the table storing
permanent identity objects (that are used as keys in the other table).
However, the old mapping was not removed while replacing the mapping in
the pseudonym table caused the old pseudonym to get destroyed. This
eventually caused crashes when a new pseudonym had the same hash value as
such a defunct entry and keys had to be compared.
Fixesstrongswan/strongswan#46.
If two CHILD_SAs with mark=%unique are created concurrently they could
otherwise end up with either the same mark or different marks in both
directions.
This is the case for the IKE_SA_INIT and the initial IKEv1 messages, which
are pre-generated in tasks as at least parts of it are used to generate
the AUTH payload. The IKE_SA_INIT message will never be fragmented, but
the IKEv1 messages might be, so we can't just call generate_message().
Fixes#1478.
Some peers send an INITIAL_CONTACT notify after they received our XAuth
username. The XAuth task waiting for the third XAuth message handles
this incorrectly and closes the IKE_SA as no configuration payloads are
contained in the message. We queue the INFORMATIONAL until the XAuth
exchange is complete to avoid this issue.
Fixes#1434.
If the af-alg plugin is enabled, but kernel support is missing, we get
an error line during startup for each probed algorithm. This is way too
verbose, so just skip probing if AF_ALG is unsupported.
The versioning scheme used by Python (PEP 440) supports the rcN suffix
but development releases have to be named devN, not drN, which are
not supported and considered legacy versions.
This provides a fix if symmetrically overlapping policies are
installed as e.g. the case in the ikev2/ip-two-pools-db scenario:
carol 10.3.0.1/32 ----- 10.3.0.0/16, 10.4.0.0/16 moon
alice 10.4.0.1/32 ----- 10.3.0.0/16, 10.4.0.0/16 moon
Among others, the following FWD policies are installed on moon:
src 10.3.0.1/32 dst 10.4.0.0/16
...
tmpl ...
src 10.4.0.0/16 dst 10.3.0.1/32
...
src 10.4.0.1/32 dst 10.3.0.0/16
...
tmpl ...
src 10.3.0.0/16 dst 10.4.0.1/32
...
Because the network prefixes are the same for all of these they all have the
same priority. Due to that it depends on the install order which policy gets
used. For instance, a packet from 10.3.0.1 to 10.4.0.1 will match the
first as well as the last policy. However, when handling the inbound
packet we have to use the first one as the packet will otherwise be
dropped due to a template mismatch. And we can't install templates with
the "outbound" FWD policies as that would prevent using different
IPsec modes or e.g. IPComp on only one of multiple SAs.
Instead we install the "outbound" FWD policies with a lower priority
than the "inbound" FWD policies so the latter are preferred. But we use
a higher priority than default drop policies would use (in case they'd
be defined with the same subnets).
After adding the read callback the state is WATCHER_QUEUED and it is
switched to WATCHER_RUNNING only later by an asynchronous job. This means
that a thread that sent a Netlink message shortly after registration
might see the state as WATCHER_QUEUED. If it then tries to read the
response and the watcher thread is quicker to actually read the message
from the socket, it could block on recv() while still holding the lock.
And the asynchronous job that actually read the message and tries to queue
it will block while trying to acquire the lock, so we'd end up in a deadlock.
This is probably mostly a problem in the unit tests.
If there is currently no route to reach the other peer we just default
to left=%any. The local address is only really used to resolve
leftsubnet=%dynamic anyway (and perhaps for MIPv6 proxy transport mode).
Fixes#1375.
Metrics are basically defined to order routes with equal prefix, so ordering
routes by metric first makes not much sense as that could prefer totally
unspecific routes over very specific ones.
For instance, the previous code did break installation of routes for
passthrough policies with two routes like these in the main routing table:
default via 192.168.2.1 dev eth0 proto static
192.168.2.0/24 dev eth0 proto kernel scope link src 192.168.2.10 metric 1
Because the default route has no metric set (0) it was used, instead of the
more specific other one, to determine src and next hop when installing a route
for a passthrough policy for 192.168.2.0/24. Therefore, the installed route
in table 220 did then incorrectly redirect all local traffic to "next hop"
192.168.2.1.
The same issue occurred when determining the source address while
installing trap policies.
Fixes 6b57790270 ("kernel-netlink: Respect kernel routing priorities for IKE routes").
Fixes#1416.