[erwinzhang7]

Fixes #4283.

The existing checks cover the shape of the rdma matrix and its null diagonal, not
whether the pairs about to be used are present, so a hostfile whose rank order places
unconnected nodes beside each other passes and then leaves every rank alone in a
group of size one, reporting success. Same class as #4210.

missing_jaccl_links reports the pairs that are used but absent: for a ring each
rank's two neighbours, for a mesh every pair. With two ranks the previous and next
neighbour are the same node, which is harmless because the pair is simply examined
twice. The error names the hosts and says that for jaccl-ring the host order is the
ring order, since that is the mistake that produces it.

Checked against a live four node mesh: the misordered hostfile is now rejected, a
ring hostfile used as a mesh is rejected, and valid mesh and valid ring hostfiles
still run at 20.87 and 16.33 GB/s respectively. Six unit tests cover mesh, ring, the two rank
edge case and a ring that is deliberately not a mesh.

[zcbenz]

Both the code comment and error message are hard to comprehend, so it checks connected nodes must have rdma set in hostfile?

[erwinzhang7]

Yes, but it checks the pairs the chosen topology will actually use, not every connected pair. For jaccl-ring that is each rank's two neighbours; for the mesh it is every pair. So a hostfile can be a valid ring and an invalid mesh, which is one of the tests.