[PATCH v2 14/28] Input: Widen pointer acceleration types to double

[Simon Thum]

On 06/09/2011 07:17 PM, Daniel Stone wrote:
> This widens almost all of the float-using code in ptrveloc.[ch] to
> doubles, other than values coming from properties which are specified to
> be floats by the property API.
> 
> Bumps input API to v14 as this changes the AccelScheme signature, as
> used by xf86-input-synaptics.
> 
> Signed-off-by: Daniel Stone <daniel at fooishbar.org>
> ---
>  dix/ptrveloc.c                 |  214 +++++++++++++++++++--------------------
>  hw/xfree86/common/xf86Module.h |    2 +-


> -    float mult = 0.0, tmpf;
> -    int dx = 0, dy = 0, tmpi;
> +    double mult = 0.0, tmpf;
> +    double dx = 0.0, dy = 0.0;
[...]

>              if (dx) {
> -                tmpf = mult * (float)dx +
> +                tmpf = mult * (double)dx +
>                         dev->last.remainder[0];
>                  tmpi = (int) tmpf;
>                  valuator_mask_set(val, 0, tmpi);
> -                dev->last.remainder[0] = tmpf - (float)tmpi;
> +                dev->last.remainder[0] = tmpf - (double)tmpi;
>              }
>              if (dy) {
I think this really should be if (dx != 0.0). If only to be explicit.
(There are 4 instances of that.)

Otherwise,

Reviewed-by: Simon Thum <simon.thum at gmx.de>

Cheers,

Simon