sbruno@ has had to revert r307901 (inform CC modules about loss events) and lstewart@ knows about this.
https://reviews.freebsd.org/D11003 still has not been committed but is now being tested at LLNW. The code cannot be MFC'd due to structure changes.
Various discussions continue on BBR. LLNW believes they have a mis-merge are and trying to work through that.
jegg@ asks about running BBR over LAGG but rrs@ says that his tests do not show a problem with LAGG and BBR. The LLNW folks will also grab a blackbox trace.
rrs@ has abstracted the send buffer code and is working on this with lstewart@