Loading
Bug 1954423: Make sure we finish checking peer nominated pairs. r=mjf
This contains three changes. 1. When nomination happens, do not cancel higher priority pairs that have been peer nominated but not fully nominated yet (because our own checks have not succeeded). If our response to the peer nomination on a pair like this has been received, the other end is already using it. 2. Allow CANCELED pairs to be revived by the triggered check code, similar to IN_PROGRESS pairs. 3. Allow the triggered check code to run again on a pair if that pair just received a peer nomination, and has a priority high enough that it could end up being selected. Differential Revision: https://phabricator.services.mozilla.com/D242720