Re: [CFT] more kdev_t-ectomy

    > Of course it may be possible to avoid kernel-internal numbers altogether.
    > Sometimes that is an improvement, sometimes not. Pointers are more
    > complicated than numbers - they point at something that must be allocated
    > and freed and reference counted. A number is like a pointer without the
    > reference counting.

    I guess the question is: is there any point to have three forms --
    with necessary conversions between them -- or is it simpler to have
    two forms and just use the more awkward dev_t form everywhere?

It doesnt matter much. I would not have introduced kdev_t just
for slightly more efficient dev_t handling. But we have it already.
It seems meaningless to go and replace it by something more awkward
and less efficient.

[But should anyone want: globally s/kdev_t/dev_t/ and a small edit
of kdev_t.h suffices.]

    We do need a dev32_t for NFSv2 et al, though.

I don't know why.

Andries