| … | |
… | |
| 2 | |
2 | |
| 3 | TODO: bridge mode, finally? |
3 | TODO: bridge mode, finally? |
| 4 | TODO: gcm mode? |
4 | TODO: gcm mode? |
| 5 | TODO: replace ripemd160 as the only authentication hash. |
5 | TODO: replace ripemd160 as the only authentication hash. |
| 6 | TODO: increase rsa size. |
6 | TODO: increase rsa size. |
|
|
7 | TODO: replace transport bits by transport endpoint structs? |
|
|
8 | TODO: ecdh to avoid session replay attacks |
|
|
9 | TODO: http://incog-izick.blogspot.de/2011/08/using-openssl-aes-gcm.html |
|
|
10 | TODO: http://stackoverflow.com/questions/12153009/openssl-c-example-of-aes-gcm-using-evp-interfaces |
| 7 | |
11 | |
|
|
12 | proposed: 3 types, req, resreq, res |
|
|
13 | |
|
|
14 | req (hmac1) rsa(seqno1 hmac1 aes1 seqno2 hmac2 aes2 auth) ecdh1 |
|
|
15 | res (hmac1) hash(rsa-contents) ecdh2 |
|
|
16 | |
|
|
17 | req hmac0(*) rsa(seqno hmac0 hmac aes auth) hkdf-salt ecdh1 |
|
|
18 | res hmac0(rsa-contents ecdh2) |
|
|
19 | hmac_key = hkdf(hkdf-salt, hmac | ecdh) |
|
|
20 | aes_key = hkdf(hkdf-salt, aes | ecdh) |
|
|
21 | |
|
|
22 | TODO: protocol magic fixen(!!!) |
|
|
23 | TODO: "global" |
|
|
24 | TODO: verify |
|
|
25 | TODO: make sense of overhead calculation |
|
|
26 | TODO: if-up &c should not be scripts? |
|
|
27 | TODO: low-power |
|
|
28 | - INCOMPATIBLE CHANGE: core protocol version 1.0. |
|
|
29 | - INCOMPATIBLE CHANGE: node sextions are now introduced |
|
|
30 | with "node nodename", not "node = nodename". |
|
|
31 | - while individual packets couldn't be replayed, a whole session |
|
|
32 | could be replayed - this has been fixed by an extra key exchange. |
|
|
33 | - in addition to rsa key exchange and authentication, the handshake now |
|
|
34 | adds a diffie-hellman key exchange (using curve25119) for perfect |
|
|
35 | forward secrecy. mac and cipher keys are derived using HKDF. |
|
|
36 | - rsa key sizes are now configurable and larger (default is 3072). |
|
|
37 | correspondingly, the minimum mtu is no longer 296 but 576. |
|
|
38 | - fixed a potential (unverified) buffer overrun on rsa decryption. |
|
|
39 | - new per-node low-power setting that tries to reduce cpu/network usage. |
|
|
40 | - router reconnects could cause excessive rekeying on other connections. |
|
|
41 | - gvpectrl no longer generates all missing public keys, but |
|
|
42 | only missing private keys. private keys are also put |
|
|
43 | into the configured location. |
|
|
44 | - the pid-file now accepts %s as nodename as elsewhere. |
|
|
45 | - no longer use RAND_bytes to generate session keys - you NEED |
|
|
46 | a real source of entropy now (e.g. egd or /dev/random - see the |
|
|
47 | openssl documentation). |
|
|
48 | - multiple node statements for the same node are now supported |
|
|
49 | and will be merged. |
|
|
50 | - a new directive "global" switches back to the global section |
|
|
51 | of the config file. |
|
|
52 | - 12 random prefix bytes are now properly supported, leading to |
|
|
53 | a fully random IV. |
|
|
54 | - use aes with a random key in counter-mode to generate IVs, |
|
|
55 | for speed reasons (generating 12 random bytes with openssl can |
|
|
56 | take longer than encrypting and mac'ing a 1.5kb packet, and IVs |
|
|
57 | do not need to be cryptographically strong random numbers |
|
|
58 | (and in fact, shouldn't be)). |
|
|
59 | - if-up scripts can now be specified with absolute paths. |
|
|
60 | - new global option: serial, to detect configuration mismatches. |
|
|
61 | - use HKDF as authentication proof, not HMAC or a plain hash |
|
|
62 | (hint by Ilmari Karonen). |
|
|
63 | - during rekeying or conenction establishments, hmac authentication |
|
|
64 | errors could occur and reset the connection. Transient hmac |
|
|
65 | authentication errors are now being ignored for 3 seconds. |
|
|
66 | - log the reason for a conneciton loss. |
|
|
67 | |
|
|
68 | 2.25 Sat Jul 13 06:42:33 CEST 2013 |
| 8 | - INCOMPATIBLE CHANGE: no longer enable udp protocol if no other |
69 | - INCOMPATIBLE CHANGE: no longer enable udp protocol if no other |
| 9 | protocols are enabled - this is necessary when you have nodes with |
70 | protocols are enabled - this is necessary when you have nodes with |
| 10 | completely unknown protocols, to force mediated connection requests. |
71 | completely unknown protocols, to force mediated connection requests. |
| 11 | - INCOMPATIBLE CHANGE: dns transport protocol bumped to version 2. |
72 | - INCOMPATIBLE CHANGE: dns transport protocol bumped to version 2. |
| 12 | - core protocol version 0.1, compatible with older releases. |
73 | - core protocol version 0.1, compatible with older releases. |
| … | |
… | |
| 20 | the protocol not being configured locally. |
81 | the protocol not being configured locally. |
| 21 | - new global configuration options: chroot, chuser, chuid, chgid, |
82 | - new global configuration options: chroot, chuser, chuid, chgid, |
| 22 | to chroot to a specified or anonymous new root, and change user id. |
83 | to chroot to a specified or anonymous new root, and change user id. |
| 23 | - new global configuration options seed_device and seed_interval, |
84 | - new global configuration options seed_device and seed_interval, |
| 24 | to configure another device than /dev/urandom for random seeds, |
85 | to configure another device than /dev/urandom for random seeds, |
| 25 | and to configure a regular imterval to reseed the rng. |
86 | and to configure a regular interval to reseed the rng. |
| 26 | - prefer inet_aton over gethostbyname, as the latter is not guaranteed |
87 | - prefer inet_aton over gethostbyname, as the latter is not guaranteed |
| 27 | to "resolve" literal ip addresses. |
88 | to "resolve" literal ip addresses. |
| 28 | - configure didn't detect openssl 1.0 because SHA1_version became private |
89 | - configure didn't detect openssl 1.0 because SHA1_version became private |
| 29 | (patch by TANIGUCHI Takaki). |
90 | (patch by TANIGUCHI Takaki). |
| 30 | - fix a bug where nodes would tell the other side that it supports |
91 | - fix a bug where nodes would tell the other side that it supports |
| … | |
… | |
| 35 | - gvpectrl no longer evaluates any "on" directives. |
96 | - gvpectrl no longer evaluates any "on" directives. |
| 36 | - icmp and rawip protocols weren't upgradable to each other. |
97 | - icmp and rawip protocols weren't upgradable to each other. |
| 37 | - major, but incremental, dns transport improvements: |
98 | - major, but incremental, dns transport improvements: |
| 38 | - do not simply abort in some error cases in the dns transport, |
99 | - do not simply abort in some error cases in the dns transport, |
| 39 | but try to recover. |
100 | but try to recover. |
| 40 | - allow lowercase/uppercase alises for base-n encodings that do |
101 | - allow lowercase/uppercase aliases for base-n encodings that do |
| 41 | not rely on case. |
102 | not rely on case. |
| 42 | - use base26 instead of base22 encoding for dns syn's, and |
103 | - use base26 instead of base22 encoding for dns syn's, and |
| 43 | base36 instead of base22 for headers (saves one byte/packet). |
104 | base36 instead of base22 for headers (saves one byte/packet). |
| 44 | - back off far quicker in dns tunnel when idling - increases |
105 | - back off far quicker in dns tunnel when idling - increases |
| 45 | latency on an idle link somewhat, but avoids hundreds of |
106 | latency on an idle link somewhat, but avoids hundreds of |
