Re: [RFC PATCH] clocksource/drivers/sh_cmt: access registers according to spec

[Wolfram Sang <wsa+renesas@sang-engineering.com>]

[-- Attachment #1: Type: text/plain, Size: 853 bytes --]


> unsigned int (everywhere)?
> Yes, this is the result of a long-by-long division, but it should be
> a small value anyway.

OK.

> > +       /* Tests showed that we need to wait 3 clocks here */
> > +       unsigned long cmcnt_delay = ch->cmt->reg_delay * 3 / 2;
> 
> DIV_ROUND_UP()

Really here? reg_delay is ensured to be a multiple of two...

> > +       /* We shall wait 2 input clks after register writes */
> > +       if (cmt->info->model >= SH_CMT_48BIT) // FIXME: Really not needed for older ones?
> > +               cmt->reg_delay = 2000000UL / rate + 1;

... but here it would make a lot of sense!

> > +       cmt->rate = cmt->info->width == 16 ? rate / 512 : rate / 8;
> 
> cmt->rate = rate / (cmt->info->width == 16 ? 512 : 8);

OK. I had it without the parens and it looked ugly. With parens it looks
good.


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]