Code review comment for lp:~kdub/mir/nested-passthrough

@Daniel
for (1) the issue is that the client is swapinterval0, and the chain representing that client in the host server is operating with 'swapinterval1'. The spare buffers quickly all end up getting stuck in the host's queue, instead of returning from the host on new submissions.

I'm plumbing up a way for the host and the client to agree on swapinterval at the moment. will restack on that branch, and then we don't have to have the issue.

for (3) can probably add this in a subsequent branch.

@Cemil
Will propose updates to the latency suite to accommodate nested passthrough, hopefully before the sprint.