Age | Commit message (Collapse) | Author | |
---|---|---|---|
2013-04-17 | Fix handling of the local address in shell commands | Matthias Schiffer | |
Without this fix, using on-establish/disestablish/verify would cause a strange zero port when a bind with a random port was used, and a segmentation fault with dynamic binds. | |||
2013-03-30 | Set development version | Matthias Schiffer | |
2013-03-21 | Don't read the default peer group twice on startup | Matthias Schiffer | |
There are no negative consequences when the peers are loaded twice, but we can spare some log spam like this. | |||
2013-03-21 | Remove debian init script again | Matthias Schiffer | |
This reverts commit 0f7aac271c521c2e30b833956eba40e42a79c420. The scripts live in the /debian directory of the Debian package source, there is no reason to duplicate them in this git. | |||
2013-03-12 | add debian init script to examples/ | Nils Schneider | |
fastd-debian-init should be copied to /etc/init.d/fastd fastd-debian-default should be copied to /etc/default/fastd | |||
2013-03-10 | fastd v8v8 | Matthias Schiffer | |
2013-03-09 | Add options to hide IP and MAC addresses from log output | Matthias Schiffer | |
2013-03-09 | Don't try to add temporary peers for disabled keys | Matthias Schiffer | |
2013-03-08 | Fix crash on invalid key definitions | Matthias Schiffer | |
2013-03-08 | Handle duplicate keys | Matthias Schiffer | |
When two peers are configured with the same key, disable both. When a temporary peer's key is configured, delete the temporary key. | |||
2013-03-08 | Really fix adding new peers from a peer dir | Matthias Schiffer | |
2013-03-08 | Don't forget all peers when no peers are configured in peer dirs | Matthias Schiffer | |
2013-03-08 | Don't lose all but one peer configured in peer dirs after loading | Matthias Schiffer | |
This regression was introduced in 4e9b2a8819c7062ad052cd6c549db1d0fbed001d. | |||
2013-03-08 | Allow disabling previously enabled peers | Matthias Schiffer | |
2013-03-08 | Disable peer configs by default, enable on peer creation | Matthias Schiffer | |
This allows to remove some duplicate code, and will simplify the detection and handling of duplicate keys. | |||
2013-03-05 | Simplify git-describe call and make it more robust | Matthias Schiffer | |
2013-03-04 | Fix git-describe for paths with spaces | Matthias Schiffer | |
2013-03-04 | Suppress error message when no .git directory is found | Matthias Schiffer | |
2013-03-04 | Fix git-describe version command | Matthias Schiffer | |
2013-03-04 | Load peer dirs in the last step of the configuration | Matthias Schiffer | |
Loading the peer dirs directly led to peers being discovered in a different order after reconfigure. | |||
2013-03-03 | Make 'peer limit 0' allow no connections | Matthias Schiffer | |
2013-03-03 | Fix typo in --log-level help text | Matthias Schiffer | |
2013-03-03 | Remove the word "version" from usage text | Matthias Schiffer | |
It is now included in the version string itself. | |||
2013-03-02 | Don't crash on empty UDP packets | Matthias Schiffer | |
2013-03-02 | Print error message on aborts due to buffer push/pull errors | Matthias Schiffer | |
2013-03-01 | Re-verify unknown peers on each handshake | Matthias Schiffer | |
2013-03-01 | Identify peers be key in log output when no name is available | Matthias Schiffer | |
2013-02-27 | Allow temporary peers when no other floating or dynamic peers are configured | Matthias Schiffer | |
2013-02-27 | Don't include linux/if_alg.h when it is not needed | Matthias Schiffer | |
2013-02-27 | Don't set the peer address for temporary peers before the session is ↵ | Matthias Schiffer | |
actually established Doing so could lead to duplicate address entries in different peers, causing very strange behaviour. Add additional parameters for the local and the peer address to fastd_shell_exec() to allow the on-verify script to use this information nevertheless. | |||
2013-02-27 | Refactor fastd_peer_claim_address | Matthias Schiffer | |
2013-02-26 | Don't reset "seen" timestamp on peer reset | Matthias Schiffer | |
2013-02-26 | Experimental support for accepting connections from unknown peers | Matthias Schiffer | |
2013-02-25 | Fix shell command exit status message, for real. | Matthias Schiffer | |
2013-02-25 | Fix shell exit status warning condition | Matthias Schiffer | |
2013-02-25 | Rewrite 'git describe' version support to update the version whenever needed | Matthias Schiffer | |
2013-02-25 | Add public keys to shell environment | Matthias Schiffer | |
2013-02-25 | Set some more environment variables for shell commands | Matthias Schiffer | |
2013-02-25 | Get rid of some duplicate code for calling shell commands | Matthias Schiffer | |
2013-02-23 | Differentiate between reasons for ignoring a handshake | Matthias Schiffer | |
2013-02-23 | Subtract splay time to key refresh interval | Matthias Schiffer | |
A random splay time of up to 5 minutes will ensure that simultaneous handshakes with many peers are desynchronized as fast as possible. | |||
2013-02-23 | Fail initialization when a default socket can't be bound | Matthias Schiffer | |
2013-02-23 | Print port for "any" address in bind log messages | Matthias Schiffer | |
2013-02-23 | Implement simple peer dump triggered by SIGUSR1 | Matthias Schiffer | |
2013-02-23 | Use git version with --dirty | Matthias Schiffer | |
2013-02-23 | Use 'git describe' output as version string when available | Matthias Schiffer | |
2013-02-23 | Use fastd_peer_t instead of fastd_peer_config_t in handshake code | Matthias Schiffer | |
Directly using the peers allows us to get rid of the inefficient get_peer() function and is necessary for adding support for unknown peers. | |||
2013-01-24 | Remove some .gitignore entries | Matthias Schiffer | |
As these are specific to my build environment, these better go to .git/info/exclude | |||
2013-01-24 | Don't reset peer address on repeated resolves | Matthias Schiffer | |
This can lead to handshakes being ignored for an indefinite time. | |||
2013-01-24 | resolver: write whole hostname instead of a pointer to the resolver pipe | Matthias Schiffer | |
Shouldn't make a difference, but feels cleaner now, and silences a clang-analyzer warning. |