Re: [patch 2.6.26-git] /dev/hpet - fixes and cleanup

[David Brownell <david-b@pacbell.net>]

On Monday 28 July 2008, Clemens Ladisch wrote:
> David Brownell wrote:
> > Oh ... I see.  It's called write_counter() but doesn't
> > actually write the counter.  Likewise, read_counter() is
> > not actually reading the counter.  Gaack ...
> 
> A more correct name would be something like
> write_a_register_that_has_the_same_size_as_a_counter_register().

writel() would be my choice ... :)

I don't think anything in that code really needs to modify the
upper 32 bits of those registers.  And nothing except the ia64
HPET clocksource needs to read them either.

 
> I'm hoping that I can deprecate and delete much of this code without
> having to clean it up first.

I'm not sure how practical that goal is, but I agree that there
seems to be no point to /dev/hpet ... certainly, you could start
by deprecating it and making sure distroes don't enable it.


> > Any comments on the rest?
> 
> Looks fine.  I would have preferred a series of smaller patches, but my
> feelings on this are not strong enough to split it myself or to ask you
> to do it.

OK, I'll send along the updated patch (against 2.6.27-rc1).

- Dave