Re: [CFT] more kdev_t-ectomy

> MKDEV(<constant>,<constant>) is a valid thing, as far as I'm concerned.

Yes. I was tempted to change the first argument of blk_register_region
into a pair, killing some MKDEV occurrences, but then I noticed that
almost all are of the form MKDEV(<constant>,<constant>), and that
is not so bad.

Still, the fact that every single call of blk_register_region
has a first argument MKDEV(ma,mi) suggests that one might
consider leaving these parameters separate.

Andries

[Now that we are talking anyway, let me ask about something.
You wrote blk_register_region so that subregions override
superregions. At the bottom there is the full region.
Was this just a general good idea, or do you have definite
applications in mind? I ask this mostly because the hash
lookup becomes more complicated in the general case.
You may have noticed that I wrote

static inline int major_to_index(int major)
{
        return major % MAX_PROBE_HASH;
}
static inline int dev_to_index(dev_t dev)
{
        return major_to_index(MAJOR(dev));
}

and that is OK for regions with constant major.
For multimajor regions a hash does not work very well, and
a tree looks better.]