Caleb James DeLisle
|
7ba35479e7
manual merge crashey
|
2 years ago |
Rogério Sampaio de Almeida
|
4d0aefc5a5
remove unused param from new timing funcs
|
2 years ago |
Caleb James DeLisle
|
027ad407c5
incomingConnection should default to false, and the version should be provided if it is known
|
3 years ago |
Caleb James DeLisle
|
1c0d1de7cb
Need to also include login for beacon replies
|
3 years ago |
Caleb James DeLisle
|
7fbb3bfa9c
Add 4 beacon tests for noise/old, noise/noise, old/noise and old/old crypto protocols
|
3 years ago |
Caleb James DeLisle
|
eb83a46260
Implemented boringtun protocol
|
3 years ago |
Caleb James DeLisle
|
4eebb33195
NOISE should forward empty packets so that InterfaceController knows that a peer has been registered
|
3 years ago |
Caleb James DeLisle
|
9a9ffd57d5
Return anyhow::Error from Iface_DEFUN functions rather than an integer error type
|
3 years ago |
Caleb James DeLisle
|
e022fa45a3
Better logging and fix what was causing noise packets to be forwarded
|
3 years ago |
Caleb James DeLisle
|
29afc1a127
Try enabling NOISE proto
|
3 years ago |
Caleb James DeLisle
|
a0084a9e21
Switch to Rust CryptoAuth
|
3 years ago |
Caleb James DeLisle
|
8e3011cb6e
Change message->bytes to message->msgbytes in prep for privatizing it
|
3 years ago |
Caleb James DeLisle
|
3651b0d413
Message privatize length
|
3 years ago |
Caleb James DeLisle
|
a73e2c68f7
privatized message.capacity
|
3 years ago |
Caleb James DeLisle
|
cd188e865c
Message: Privatized alloc
|
3 years ago |
Caleb James DeLisle
|
7272cbdfc9
Got the new cryptoauth working with message-passthrough model
|
3 years ago |
Caleb James DeLisle
|
ccf923d8f6
Major refactor, switch CryptoAuth to be an Iface adaptor rather than using encrypt() and decrypt() functions
|
3 years ago |
Caleb James DeLisle
|
6a1a5ea073
Wrong type for user in newSession
|
3 years ago |
Caleb James DeLisle
|
6d9ecb7e55
re-shuffled tests to be able to run on both new and old CryptoAuth
|
3 years ago |
Caleb James DeLisle
|
3cf79da359
Iface_DEFUN return an error type
|
4 years ago |
Caleb James DeLisle
|
d8060ad7f5
Don't announce peers to PF if they're an incompatible protocol
|
4 years ago |
Caleb James DeLisle
|
f40430faf4
memory leak
|
4 years ago |
Caleb James DeLisle
|
78da0951b5
1. Present incompatible peers as INCOMPATIBLE rather than sticking on ESTABLISHED
|
4 years ago |
Caleb James DeLisle
|
1477c96bad
Oops, meant to use a return statment there
|
4 years ago |
Caleb James DeLisle
|
286d751b6e
Reduce logging noise
|
4 years ago |
Caleb James DeLisle
|
40eb3b402c
SUBNODE: Better solution: At the InterfaceController level do not report v20 or older peers
|
4 years ago |
Caleb James DeLisle
|
2c98115d5c
When in SUBNODE mode, refuse to talk to version <21
|
4 years ago |
Caleb James DeLisle
|
de159175b7
Better logging + Subnode: always ping a peer when it emerges
|
4 years ago |
Caleb James DeLisle
|
63c60852f7
Minor logging improvements
|
4 years ago |
Caleb James DeLisle
|
ddd2161302
InterfaceController: do not move endpoints because it is possible to have multiple valid peerings with the same node
|
4 years ago |