Commit Graph

6 Commits (18c61afeb9ebec3169aab0e2fa75c996de84a250)

Author SHA1 Message Date
Tom DNetto a515fc517b ipn/ipnlocal: make tkaSyncIfNeeded exclusive with a mutex
Running corp/ipn#TestNetworkLockE2E has a 1/300 chance of failing, and
deskchecking suggests thats whats happening are two netmaps are racing each
other to be processed through tkaSyncIfNeededLocked. This happens in the
first place because we release b.mu during network RPCs.

To fix this, we make the tka sync logic an exclusive section, so two
netmaps will need to wait for tka sync to complete serially (which is what
we would want anyway, as the second run through probably wont need to
sync).

Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Tom DNetto 8602061f32 ipn/ipnlocal,tka: Fix bugs found by integration testing
* tka.State.staticValidateCheckpoint could call methods on a contained key prior to calling StaticValidate on that key
 * Remove broken backoff / RPC retry logic from tka methods in ipn/ipnlocal, to be fixed at a later time
 * Fix NetworkLockModify() which would attempt to take b.mu twice and deadlock, remove now-unused dependence on netmap
 * Add methods on ipnlocal.LocalBackend to be used in integration tests
 * Use TAILSCALE_USE_WIP_CODE as the feature flag so it can be manipulated in tests

Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Tom DNetto 73db56af52 ipn/ipnlocal: filter peers with bad signatures when tka is enabled
Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Tom DNetto 58ffe928af ipn/ipnlocal, tka: Implement TKA synchronization with the control plane
Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Tom DNetto ebd1637e50 ipn/ipnlocal,tailcfg: Identify client using NodeKey in tka RPCs
Updates https://github.com/tailscale/corp/pull/7024

Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago
Tom DNetto e9b98dd2e1 control/controlclient,ipn/ipnlocal: wire tka enable/disable
Signed-off-by: Tom DNetto <tom@tailscale.com>
2 years ago