Re: [PATCH] qemu_command: Generate memory only after controllers

I was imagining some change happening in the future under the assumption that memory devices will only ever be plugged into pci(e).0: the current test suite would not catch that, but if we had at least one case in which a memory device is plugged into a different bus then we'd have a chance of noticing. Maybe that's a bit far-fetched, but even though the potential gain would be small the effort needed to obtain it would also be small (something along the lines of the diff below), so overall it seems like a decent deal to me :)