[v2,4/4] pwm: rcar: improve calculation of divider

The rcar_pwm_get_clock_division() has a loop to calculate the divider,
but the value of div should be calculatable without a loop. So,
this patch improves it.

This algorithm is suggested by Uwe Kleine-König and Laurent Pinchart.

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
---
 drivers/pwm/pwm-rcar.c | 16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)

On Tue, Jan 08, 2019 at 12:28:14PM +0900, Yoshihiro Shimoda wrote:
> The rcar_pwm_get_clock_division() has a loop to calculate the divider,
> but the value of div should be calculatable without a loop. So,
> this patch improves it.
> 
> This algorithm is suggested by Uwe Kleine-König and Laurent Pinchart.
> 
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>

Best regards
Uwe

Hi Shimoda-san,

On Tue, Jan 8, 2019 at 4:31 AM Yoshihiro Shimoda
<yoshihiro.shimoda.uh@renesas.com> wrote:
> The rcar_pwm_get_clock_division() has a loop to calculate the divider,
> but the value of div should be calculatable without a loop. So,
> this patch improves it.
>
> This algorithm is suggested by Uwe Kleine-König and Laurent Pinchart.
>
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> ---
>  drivers/pwm/pwm-rcar.c | 16 +++++++---------
>  1 file changed, 7 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/pwm/pwm-rcar.c b/drivers/pwm/pwm-rcar.c
> index 6dbb70c..0498a93 100644
> --- a/drivers/pwm/pwm-rcar.c
> +++ b/drivers/pwm/pwm-rcar.c
> @@ -8,6 +8,8 @@
>  #include <linux/clk.h>
>  #include <linux/err.h>
>  #include <linux/io.h>
> +#include <linux/log2.h>
> +#include <linux/math64.h>
>  #include <linux/module.h>
>  #include <linux/of.h>
>  #include <linux/platform_device.h>
> @@ -68,19 +70,15 @@ static void rcar_pwm_update(struct rcar_pwm_chip *rp, u32 mask, u32 data,
>  static int rcar_pwm_get_clock_division(struct rcar_pwm_chip *rp, int period_ns)
>  {
>         unsigned long clk_rate = clk_get_rate(rp->clk);
> -       unsigned long long max; /* max cycle / nanoseconds */
> -       unsigned int div;
> +       u64 div, tmp;
>
>         if (clk_rate == 0)
>                 return -EINVAL;
>
> -       for (div = 0; div <= RCAR_PWM_MAX_DIVISION; div++) {
> -               max = (unsigned long long)NSEC_PER_SEC * RCAR_PWM_MAX_CYCLE *
> -                       (1 << div);
> -               do_div(max, clk_rate);
> -               if (period_ns <= max)
> -                       break;
> -       }
> +       div = NSEC_PER_SEC * RCAR_PWM_MAX_CYCLE;

As we have:

    #define NSEC_PER_SEC    1000000000L
    #define RCAR_PWM_MAX_CYCLE      1023

NSEC_PER_SEC is 64-bit on arm64, and 32-bit on arm32.

Hence you should use

    div = (u64)NSEC_PER_SEC * RCAR_PWM_MAX_CYCLE;

to avoid overflow on arm32.

> +       tmp = (u64)period_ns * clk_rate + div - 1;
> +       tmp = div64_u64(tmp, div);
> +       div = ilog2(tmp - 1) + 1;
>
>         return (div <= RCAR_PWM_MAX_DIVISION) ? div : -ERANGE;
>  }

Gr{oetje,eeting}s,

                        Geert

Hi Geert-san,

> From: Geert Uytterhoeven, Sent: Tuesday, January 8, 2019 5:22 PM
> 
> Hi Shimoda-san,
> 
> On Tue, Jan 8, 2019 at 4:31 AM Yoshihiro Shimoda
> <yoshihiro.shimoda.uh@renesas.com> wrote:
> > The rcar_pwm_get_clock_division() has a loop to calculate the divider,
> > but the value of div should be calculatable without a loop. So,
> > this patch improves it.
> >
> > This algorithm is suggested by Uwe Kleine-König and Laurent Pinchart.
> >
> > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> > ---
> >  drivers/pwm/pwm-rcar.c | 16 +++++++---------
> >  1 file changed, 7 insertions(+), 9 deletions(-)
> >
> > diff --git a/drivers/pwm/pwm-rcar.c b/drivers/pwm/pwm-rcar.c
> > index 6dbb70c..0498a93 100644
> > --- a/drivers/pwm/pwm-rcar.c
> > +++ b/drivers/pwm/pwm-rcar.c
> > @@ -8,6 +8,8 @@
> >  #include <linux/clk.h>
> >  #include <linux/err.h>
> >  #include <linux/io.h>
> > +#include <linux/log2.h>
> > +#include <linux/math64.h>
> >  #include <linux/module.h>
> >  #include <linux/of.h>
> >  #include <linux/platform_device.h>
> > @@ -68,19 +70,15 @@ static void rcar_pwm_update(struct rcar_pwm_chip *rp, u32 mask, u32 data,
> >  static int rcar_pwm_get_clock_division(struct rcar_pwm_chip *rp, int period_ns)
> >  {
> >         unsigned long clk_rate = clk_get_rate(rp->clk);
> > -       unsigned long long max; /* max cycle / nanoseconds */
> > -       unsigned int div;
> > +       u64 div, tmp;
> >
> >         if (clk_rate == 0)
> >                 return -EINVAL;
> >
> > -       for (div = 0; div <= RCAR_PWM_MAX_DIVISION; div++) {
> > -               max = (unsigned long long)NSEC_PER_SEC * RCAR_PWM_MAX_CYCLE *
> > -                       (1 << div);
> > -               do_div(max, clk_rate);
> > -               if (period_ns <= max)
> > -                       break;
> > -       }
> > +       div = NSEC_PER_SEC * RCAR_PWM_MAX_CYCLE;
> 
> As we have:
> 
>     #define NSEC_PER_SEC    1000000000L
>     #define RCAR_PWM_MAX_CYCLE      1023
> 
> NSEC_PER_SEC is 64-bit on arm64, and 32-bit on arm32.
> 
> Hence you should use
> 
>     div = (u64)NSEC_PER_SEC * RCAR_PWM_MAX_CYCLE;
> 
> to avoid overflow on arm32.

Thank you for the comment! I'll fix this.

Best regards,
Yoshihiro Shimoda

> > +       tmp = (u64)period_ns * clk_rate + div - 1;
> > +       tmp = div64_u64(tmp, div);
> > +       div = ilog2(tmp - 1) + 1;
> >
> >         return (div <= RCAR_PWM_MAX_DIVISION) ? div : -ERANGE;
> >  }
> 
> Gr{oetje,eeting}s,
> 
>                         Geert
> 
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
> 
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds