>  > This strikes me as an easy way now to debug various problems with
>  > devices with blocksize != DEV_BSIZE...
> I found at least one problem in all of the other disk drivers (see my
> post to tech-kern regarding foosize()), and plan to address them
> shortly.

Ah, now if only disk drive manufacturers hadn't surrendered to the
incompetance of CERTAIN PC operating system vendors and given up on making
it possible to reformat with larger sectors to increase capacity...
(though IBM's "sector per track" drives dodge that problem pretty gracefully
by simply *pretending* to have 512 byte sectors...)