[RFC V1 3/8] clk: Add support for simple dividers
Sascha Hauer
s.hauer at pengutronix.de
Thu Dec 1 09:24:14 EST 2011
On Thu, Dec 01, 2011 at 03:42:59PM +0800, Richard Zhao wrote:
> On Wed, Nov 30, 2011 at 12:43:10PM -0800, Mike Turquette wrote:
> > On Wed, Nov 23, 2011 at 3:12 AM, Richard Zhao <richard.zhao at linaro.org> wrote:
> > > +#define account_for_rounding_errors 1
> > > +
> > > +static int clk_divider_bestdiv(struct clk *clk, unsigned long rate,
> > > + unsigned long *best_parent_rate)
> > > +{
> > > + struct clk_divider *divider = to_clk_divider(clk);
> > > + int i, bestdiv = 0;
> > > + unsigned long parent_rate, best = 0, now, maxdiv;
> > > +
> > > + maxdiv = (1 << divider->width);
> > > +
> > > + if (divider->flags & CLK_DIVIDER_FLAG_ONE_BASED)
> > > + maxdiv--;
> > > +
> > > + if (!(clk->flags & CLK_PARENT_SET_RATE)) {
> > > + parent_rate = clk->parent->rate;
> > > + bestdiv = parent_rate / rate;
> > > + bestdiv = bestdiv == 0 ? 1 : bestdiv;
> > > + bestdiv = bestdiv > maxdiv ? maxdiv : bestdiv;
> > > + goto out;
> > > + }
> The code is originally from Sascha. I think it again.
>
> Sascha, why don't we use the below code?
>
> for (i = 1; i <= maxdiv; i++) {
> int div;
> parent_rate = clk_round_rate(clk->parent, rate * i);
> div = parent_rate / rate;
> div = div > maxdiv ? maxdiv : div;
> div = div < 1 ? 1 : div;
> now = parent_rate / div;
>
> if (now <= rate && now >= best) {
> bestdiv = div;
> best = now;
> best_parent_rate = parent_rate;
> }
> }
Yes, why not ;)
I can't remember the exact problem, but it came in with cascaded
dividers which wouldn't work without this. I've never been very
proud of this account_for_rounding_errors, so feel free to remove
it and see what happens. If this becomes a problem we can search
for a better solution.
>
>
> > > +
> > > + /*
> > > + * The maximum divider we can use without overflowing
> > > + * unsigned long in rate * i below
> > > + */
> > > + maxdiv = min(ULONG_MAX / rate, maxdiv);
> > > +
> > > + for (i = 1; i <= maxdiv; i++) {
> > > + parent_rate = clk_round_rate(clk->parent,
> > > + (rate + account_for_rounding_errors) * i);
> >
> > How about just ((rate + 1) * i) with a comment above explaining why?
> > This removes the awkward #define above.
> >
> > Also, I'd like to roll this into clk-basic.c for the V4 series. Any objections?
> Go ahead.
>
> Thanks
> Richard
> >
> > Thanks,
> > Mike
> >
> > _______________________________________________
> > linux-arm-kernel mailing list
> > linux-arm-kernel at lists.infradead.org
> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> >
>
>
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
More information about the linux-arm-kernel
mailing list