On Thu, Nov 05, 2020 at 04:41:46PM +0000, Paul Barker wrote:
> To convert the number of pulses counted into an RPM estimation, we need
> to divide by the width of our measurement interval instead of
> multiplying by it.
>
> We also don't need to do 64-bit division, with 32-bits we can handle a
> fan running at over 4 million RPM.
>
> As the sample_timer() function is called once per second, delta will usually
> be just over 1000 and should never be zero, avoiding the risk of a divide by
> zero exception.
>
> Signed-off-by: Paul Barker <pbarker@xxxxxxxxxxxx>
> ---
> drivers/hwmon/pwm-fan.c | 7 +++----
> 1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
> index bdba2143021a..2f2950629892 100644
> --- a/drivers/hwmon/pwm-fan.c
> +++ b/drivers/hwmon/pwm-fan.c
> @@ -54,14 +54,13 @@ static irqreturn_t pulse_handler(int irq, void *dev_id)
> static void sample_timer(struct timer_list *t)
> {
> struct pwm_fan_ctx *ctx = from_timer(ctx, t, rpm_timer);
> + unsigned int delta = ktime_ms_delta(ktime_get(), ctx->sample_start);
> int pulses;
> - u64 tmp;
>
> pulses = atomic_read(&ctx->pulses);
> atomic_sub(pulses, &ctx->pulses);
> - tmp = (u64)pulses * ktime_ms_delta(ktime_get(), ctx->sample_start) * 60;
> - do_div(tmp, ctx->pulses_per_revolution * 1000);
> - ctx->rpm = tmp;
> + ctx->rpm = (unsigned int)(pulses * 1000 * 60) /
> + (ctx->pulses_per_revolution * delta);
delta can, at least in theory, be 0. While that is quite unlikely to happen,
the code should handle this situation.
>
> ctx->sample_start = ktime_get();
> mod_timer(&ctx->rpm_timer, jiffies + HZ);
