]>
git.ipfire.org Git - thirdparty/strongswan.git/log
Martin Willi [Sat, 11 May 2013 07:58:51 +0000 (09:58 +0200)]
git-issue: get "fixed" status for version having a "issue-fixed" commit
Martin Willi [Sat, 11 May 2013 07:45:37 +0000 (09:45 +0200)]
git-issue: define issue-cve and issue-fixed keywords
Martin Willi [Sat, 11 May 2013 07:34:39 +0000 (09:34 +0200)]
issue: identification snprintf() overflow vulnerability
issue-status: fixed
issue-type: security
issue-cve: CVE-2010-2628
issue-keyword: printf, overflow
issue-affected:
8309798f
issue-fixed:
b8aeaea8
Martin Willi [Sat, 11 May 2013 07:22:02 +0000 (09:22 +0200)]
issue: GMP RSA signature verification bypass
issue-status: fixed
issue-type: security
issue-cve: CVE-2012-2388
issue-keyword: gmp
issue-affected:
552cc11b
issue-fixed:
060b508e 30004b26
Martin Willi [Sat, 11 May 2013 07:16:24 +0000 (09:16 +0200)]
issue: OpenSSL ECDSA certificate signature verification bypass
issue-status: fixed
issue-type: security
issue-cve: CVE-2013-2944
issue-keyword: openssl
issue-affected:
472cb4ce
issue-fixed:
0faaab20
Martin Willi [Fri, 10 May 2013 17:22:59 +0000 (19:22 +0200)]
git-issue: respect issue-affected when checking issue status
Martin Willi [Fri, 10 May 2013 17:00:28 +0000 (19:00 +0200)]
git-issue: some cleanups, use $(...) command substitution instead of backquotes
Martin Willi [Fri, 10 May 2013 16:50:05 +0000 (18:50 +0200)]
git-issue: speed up issue status lookup
Martin Willi [Fri, 10 May 2013 16:30:23 +0000 (18:30 +0200)]
git-issue: fix and speed up issue type lookup
Martin Willi [Fri, 10 May 2013 16:07:50 +0000 (18:07 +0200)]
git-issue: fix verification of given issue-id
Martin Willi [Fri, 10 May 2013 11:09:39 +0000 (13:09 +0200)]
git-issue: add notes about installation of git-issue helper
Martin Willi [Fri, 10 May 2013 11:06:07 +0000 (13:06 +0200)]
git-issue: use bash for git-issue, and fix shebang
Martin Willi [Fri, 10 May 2013 10:58:18 +0000 (12:58 +0200)]
git-issue: add an info command to show status of a single issue
Martin Willi [Fri, 10 May 2013 09:36:36 +0000 (11:36 +0200)]
git-issue: list can now check issue status against a specific revision
Martin Willi [Thu, 9 May 2013 18:56:01 +0000 (20:56 +0200)]
Add an introduction to issue tracking with git commits
issue-id:
5d46e95
Martin Willi [Thu, 9 May 2013 18:07:09 +0000 (20:07 +0200)]
issue: Add an initial helper tool to track issues embedded in git commits
issue-status: fixed
issue-type: feature
issue-keyword: git-issue
Martin Willi [Thu, 9 May 2013 14:41:16 +0000 (16:41 +0200)]
Use ref_get() to make sure CHILD_SA reqids are unique
issue-id:
58075c11
issue-status: fixed
Martin Willi [Thu, 9 May 2013 14:39:12 +0000 (16:39 +0200)]
Use ref_get() to make sure IKE_SA unique IDs are unique
issue-id:
58075c11
Martin Willi [Wed, 8 May 2013 13:07:09 +0000 (15:07 +0200)]
utils: ref_get() returns the new value of the reference counter
This allows us to use ref_get() for getting unique values.
Martin Willi [Thu, 9 May 2013 14:29:18 +0000 (16:29 +0200)]
issue: CHILD_SA reqid / IKE_SA unique ID counters get accessed non-atomic
This may result in non-unique identifiers if threads access the variable
concurrently. While it is not very likely, we should fix that.
issue-status: open
issue-type: minor
issue-keyword: race, libcharon
Tobias Brunner [Tue, 7 May 2013 08:19:37 +0000 (10:19 +0200)]
testing: Don't run tests when building tkm
The problem with XML/Ada described in
9c2aba27 actually occurs when
running the tests here.
Really fixes #336.
Tobias Brunner [Mon, 6 May 2013 15:21:30 +0000 (17:21 +0200)]
testing: Don't run tests when building tkm-rpc
There are issues with some versions of the XML/Ada library on i386,
blocking the build of the testing environment when these tests are run.
TKM tests won't work in such a case but at least make-testing does not
block with this patch.
Fixes #336.
Martin Willi [Mon, 6 May 2013 15:04:36 +0000 (17:04 +0200)]
Merge branch 'tun-vip'
Beside some OS X love, this merge introduces virtual IP and route installation
support on the pfkey/pfroute kernel interfaces.
Each virtual IP gets installed on a dedicated TUN device. As Linux-like source
routes are not supported, routes for the negotiated traffic selectors get
installed using the TUN device.
To prevent IKE packets from using those routes, special exclude routes get
installed to the IKE gateway. This works for most road-warrior deployments, but
certainly does not for some more exotic configurations, such as those using
virtual-IP-to-host. Mobility is not yet supported, either.
Martin Willi [Mon, 6 May 2013 14:40:19 +0000 (16:40 +0200)]
kernel-pfroute: allow only one thread to do a route look up simultaneously
Otherwise we mess up the sequence number another thread is waiting for.
Martin Willi [Sun, 21 Apr 2013 15:05:08 +0000 (17:05 +0200)]
kernel-interface: query SAD for last use time if SPD query didn't yield one
Martin Willi [Sun, 21 Apr 2013 14:50:17 +0000 (16:50 +0200)]
child-sa: query SAD/SPD just for what we actually need to update statistics
Martin Willi [Sat, 20 Apr 2013 18:54:03 +0000 (20:54 +0200)]
kernel-pfkey: be less verbose about unexpected sequence numbers
Martin Willi [Sat, 20 Apr 2013 11:29:20 +0000 (13:29 +0200)]
kernel-pfkey: install exclude routes if kernel-net requires them
Martin Willi [Sat, 20 Apr 2013 10:28:05 +0000 (12:28 +0200)]
kernel-pfroute: add a feature flag requesting "exclude" routes
If routes installed along with policies covering the peer address affect local
IKE/ESP packets, they won't get routed correctly. To work around this issue,
the kernel interface can install "exclude" routes for the IKE peer. Not all
networking backends require this workaround, hence we export a flag for it
if it is required.
Martin Willi [Fri, 19 Apr 2013 14:58:06 +0000 (16:58 +0200)]
kernel-pfroute: remove unused interface address refcounting
Martin Willi [Fri, 19 Apr 2013 14:55:38 +0000 (16:55 +0200)]
kernel-pfroute: mark IPs installed on tun device as virtual
Martin Willi [Fri, 19 Apr 2013 13:53:45 +0000 (15:53 +0200)]
kernel-pfroute: install virtual IPs using dedicated tun devices
Martin Willi [Fri, 19 Apr 2013 12:27:31 +0000 (14:27 +0200)]
kernel-pfkey: when installing a route for a virtual IP, use its interface
When installing a route over a tun device for a virtual IP, the route must
be set over the tun, not the IKE interface.
Martin Willi [Fri, 19 Apr 2013 12:22:45 +0000 (14:22 +0200)]
kernel-interface: get_address_by_ts() can tell if a returned IP is virtual
Martin Willi [Fri, 19 Apr 2013 12:52:29 +0000 (14:52 +0200)]
kernel-interface: support enumeration of virtual-only IPs
Martin Willi [Fri, 19 Apr 2013 12:17:22 +0000 (14:17 +0200)]
kernel-pfkey: refactor route installation to a dedicate function
Martin Willi [Fri, 19 Apr 2013 10:17:25 +0000 (12:17 +0200)]
kernel-pfroute: split /0 routes to avoid conflict with default route
Martin Willi [Fri, 19 Apr 2013 10:16:12 +0000 (12:16 +0200)]
kernel-pfkey: check if we have a gateway before comparing them
Martin Willi [Fri, 19 Apr 2013 08:42:23 +0000 (10:42 +0200)]
kernel-pfkey: install route along with input, not forward policies
As forwarding policies are not available on all systems (OS X), using the
forward policy to attach the route is a bad pick. Using input policies allows
OS X to install routes.
Martin Willi [Fri, 19 Apr 2013 08:47:34 +0000 (10:47 +0200)]
kernel-pfroute: rescan address list for an interface if its state changes
It seems that we don't get address notifications if the interface is down
on OS X.
Martin Willi [Fri, 19 Apr 2013 08:46:04 +0000 (10:46 +0200)]
kernel-pfroute: add newly appearing interfaces to the interface cache
Martin Willi [Thu, 18 Apr 2013 16:26:49 +0000 (18:26 +0200)]
kernel-pfroute: implement get_nexthop()
Martin Willi [Thu, 18 Apr 2013 16:19:24 +0000 (18:19 +0200)]
kernel-pfroute: install and uninstall routes
Martin Willi [Thu, 18 Apr 2013 16:15:25 +0000 (18:15 +0200)]
kernel-pfroute: collect replies received for our own queries
Martin Willi [Thu, 18 Apr 2013 15:58:48 +0000 (17:58 +0200)]
kernel-pfroute: refactor PF_ROUTE message processing, use an enumerator
Martin Willi [Thu, 18 Apr 2013 14:53:35 +0000 (16:53 +0200)]
kernel-pfkey: use an int to set esp_port with a sysctl on OS X
Martin Willi [Thu, 18 Apr 2013 14:20:11 +0000 (16:20 +0200)]
kernel-pfroute: use INIT() macro for allocations
Martin Willi [Thu, 18 Apr 2013 14:08:12 +0000 (16:08 +0200)]
kernel-pfroute: use only a single PF_ROUTE socket for both events and queries
Martin Willi [Wed, 17 Apr 2013 12:48:58 +0000 (14:48 +0200)]
kernel-pfroute: fix length check when receiving PF_ROUTE messages
Martin Willi [Mon, 6 May 2013 13:30:07 +0000 (15:30 +0200)]
kernel-pfkey: remove obsolete pluto specific behavior
Martin Willi [Fri, 12 Apr 2013 09:32:05 +0000 (11:32 +0200)]
kernel-netlink: remove obsolete pluto specific behavior
Martin Willi [Fri, 12 Apr 2013 11:03:21 +0000 (13:03 +0200)]
tun_device: add a getter for the address previously passed to set_address()
Martin Willi [Fri, 12 Apr 2013 10:42:14 +0000 (12:42 +0200)]
tun_device: add a getter for the underlying file descriptor
Martin Willi [Tue, 16 Apr 2013 10:09:18 +0000 (12:09 +0200)]
tun-device: use host_create_netmask() to calculate interface netmask
Martin Willi [Tue, 16 Apr 2013 10:08:38 +0000 (12:08 +0200)]
host: add a netmask constructor taking the number of network bits
Martin Willi [Mon, 15 Apr 2013 14:27:31 +0000 (16:27 +0200)]
host: remove unused host_t.get_differences() method
Martin Willi [Thu, 11 Apr 2013 17:51:45 +0000 (19:51 +0200)]
host: print %#H format specifiers not as %any, but with the port
Martin Willi [Wed, 20 Mar 2013 17:15:26 +0000 (18:15 +0100)]
host: initialize sockaddr->sa_len if it is available
Martin Willi [Fri, 12 Apr 2013 11:19:59 +0000 (13:19 +0200)]
child-sa: pass traffic selector to add_sa() regardless of IPsec mode
This lets the kernel backend decide what to do with it, and in fact all kernel
interfaces already handle this correctly.
Martin Willi [Thu, 11 Apr 2013 18:22:06 +0000 (20:22 +0200)]
socket-default: to bind to one dynamic port on OS X, create v4 socket before v6
It seems that the order of binding sockets of different address families to the
same dynamic port must be v6-before-v4 on Linux, but v4-before-v6 on OS X.
Martin Willi [Thu, 11 Apr 2013 18:19:57 +0000 (20:19 +0200)]
socket-default: refactor socket pair opening to a function
Martin Willi [Thu, 11 Apr 2013 17:47:28 +0000 (19:47 +0200)]
socket-default: Don't try to send packet if we haven't a socket for given family
Martin Willi [Thu, 11 Apr 2013 17:40:31 +0000 (19:40 +0200)]
socket-default: Use -1 if socket is not available, as 0 is actually a valid fd
Martin Willi [Thu, 11 Apr 2013 17:20:33 +0000 (19:20 +0200)]
semaphore: similar to thread_create(), semaphore_create() is used by Mach
The compiler spits no warning, but the wrong symbol is used when calling
semaphore_create() from strongSwan. Override the name with a #define to force
the use of our semaphore_create().
Martin Willi [Thu, 11 Apr 2013 15:34:44 +0000 (17:34 +0200)]
.gitignore: add OS X .DS_Store files
Martin Willi [Mon, 6 May 2013 14:01:25 +0000 (16:01 +0200)]
Merge branch 'charon-cmd'
Introduce a simple IKEv1/IKEv2 command line client, charon-cmd. It does
not need any configuration files at all, but takes a few command line arguments
to establish connections as a road warrior.
Martin Willi [Mon, 22 Apr 2013 10:42:01 +0000 (12:42 +0200)]
charon-cmd: add an option to set a different server identity
Martin Willi [Mon, 22 Apr 2013 10:33:52 +0000 (12:33 +0200)]
proposals: try next if IKEv2 algorithm could not be mapped to IKEv1
Martin Willi [Mon, 22 Apr 2013 10:33:10 +0000 (12:33 +0200)]
charon-cmd: add support for different IKEv1/IKEv2 authentication profiles
Martin Willi [Mon, 22 Apr 2013 09:20:58 +0000 (11:20 +0200)]
charon-cmd: support multi-line help text for each option in usage
Martin Willi [Mon, 22 Apr 2013 08:52:17 +0000 (10:52 +0200)]
charon-cmd: add --local/remote-ts options to set traffic selectors
Martin Willi [Thu, 11 Apr 2013 17:27:27 +0000 (19:27 +0200)]
charon-cmd: Use dynamic ports with the socket-default plugin
Martin Willi [Thu, 11 Apr 2013 17:23:29 +0000 (19:23 +0200)]
settings: Add a set_default_str() to set a different default for a key
The value is set only if it is not configured in strongswan.conf or has
not been set() otherwise.
Martin Willi [Tue, 26 Mar 2013 17:04:27 +0000 (18:04 +0100)]
charon-cmd: prompt for EAP passwords on-demand using a callback credential set
Martin Willi [Tue, 26 Mar 2013 16:47:28 +0000 (17:47 +0100)]
charon-cmd: authenticate with EAP if no private key is given
Martin Willi [Tue, 26 Mar 2013 16:38:02 +0000 (17:38 +0100)]
charon-cmd: pass arguments to all handlers, even if already handled
Martin Willi [Tue, 26 Mar 2013 16:36:31 +0000 (17:36 +0100)]
charon-cmd: build long_opts only once when reading arguments in charon-cmd
Martin Willi [Tue, 26 Mar 2013 16:28:40 +0000 (17:28 +0100)]
charon-cmd: add support for using dynamic ports in charon-cmd
Martin Willi [Tue, 26 Mar 2013 16:23:38 +0000 (17:23 +0100)]
socket-dynamic: when sending from port zero, allocate a free port dynamically
Martin Willi [Tue, 26 Mar 2013 15:19:00 +0000 (16:19 +0100)]
controller: clean up job data if a thread gets cancelled waiting in a function
Controller functions are thread cancellation points, so register a cancellation
handler cleaning up job data.
Martin Willi [Tue, 26 Mar 2013 14:42:06 +0000 (15:42 +0100)]
charon-cmd: load certificates and RSA private keys
Martin Willi [Tue, 26 Mar 2013 14:07:15 +0000 (15:07 +0100)]
charon-cmd: add a connection object and its initiation to charon-cmd
Martin Willi [Tue, 26 Mar 2013 13:10:00 +0000 (14:10 +0100)]
charon-cmd: move command line options to separate file, obsolete short options
Martin Willi [Tue, 26 Mar 2013 10:26:22 +0000 (11:26 +0100)]
charon-cmd: add a stub for charon-cmd, a simple command line IKE client
Martin Willi [Mon, 6 May 2013 13:15:58 +0000 (15:15 +0200)]
Merge branch 'leak-detective-dlsym'
Replaces the use of deprecated glibc malloc hooks by overloading malloc calls
and find the originals using dlsym(RTLD_NEXT).
On OS X, we now support the use of leak detective by changing the default
malloc zone functions, backtraces get resolved using atos.
It seems that the performance bottleneck now is stack unwinding. Unfortunately
a new libunwind based backtrace() is not much faster than the libc variant; we
keep that option nonetheless for platforms without backtrace().
Martin Willi [Fri, 5 Apr 2013 14:24:38 +0000 (16:24 +0200)]
scripts: add a simple test utility to do some malloc() benchmarking/profiling
Martin Willi [Wed, 17 Apr 2013 15:45:25 +0000 (17:45 +0200)]
backtrace: use atos instead of addr2line on OS X to resolve source lines
Martin Willi [Fri, 5 Apr 2013 14:26:27 +0000 (16:26 +0200)]
backtrace: add an alternative stack unwinding implementation using libunwind
Martin Willi [Thu, 18 Apr 2013 11:07:27 +0000 (13:07 +0200)]
leak-detective: add support for OS X by hooking default malloc zone
Martin Willi [Tue, 2 Apr 2013 16:27:12 +0000 (18:27 +0200)]
leak-detective: remove unused malloc call counters
Martin Willi [Tue, 2 Apr 2013 15:41:04 +0000 (17:41 +0200)]
leak-detective: align allocations on both 32 and 64-bit systems to 32 bytes
Martin Willi [Tue, 2 Apr 2013 13:47:55 +0000 (15:47 +0200)]
leak-detective: call tzset() explicitly before enabling leak detective
tzset() is hard to whitelist on some systems, as there is no symbol involved.
Call tzset() explicitly before initialization to avoid false positives.
Martin Willi [Tue, 2 Apr 2013 11:37:06 +0000 (13:37 +0200)]
leak-detective: override malloc functions instead of using deprecated hooks
malloc hooks have become deprecated, and their use has always been problematic,
especially in multi-threaded applications. Replace the functionality by
overriding all malloc functions and query the system allocator functions
using dlsym() with RTLD_NEXT.
Martin Willi [Mon, 6 May 2013 13:05:40 +0000 (15:05 +0200)]
Merge branch 'silent-automake'
Adds support for --enable/disable-silent-rules when building with Automake
1.11 or 1.12. make V=0 or V=1 can be used to build with a different verbosity
than ./configured.
Martin Willi [Wed, 10 Apr 2013 13:36:14 +0000 (15:36 +0200)]
Add -s to make if --enable-silent-rules is ./configured
Martin Willi [Wed, 10 Apr 2013 13:35:55 +0000 (15:35 +0200)]
Use the GEN silent rule when generating files with sed
Martin Willi [Wed, 10 Apr 2013 12:49:29 +0000 (14:49 +0200)]
Use the GEN silent rule when generating oid database with perl
Martin Willi [Wed, 10 Apr 2013 12:02:53 +0000 (14:02 +0200)]
Use the GEN silent rule when generating gperf files
Martin Willi [Wed, 10 Apr 2013 12:01:41 +0000 (14:01 +0200)]
Provide the --enable-silent-rules ./configure option with automake 1.11 and 1.12
Martin Willi [Fri, 5 Apr 2013 12:10:00 +0000 (14:10 +0200)]
Introduce an optional logger_t.vlog() method with format string and arguments
To have more flexibility in the logging backend, receiving the original format
string and do printf() substitution in the logger may be preferable. An
additional but optional logger method does not touch the behavior of existing
loggers.