[PATCH v2] dix: undo transformation for missing valuators (#49347)

Chase Douglas chase.douglas at canonical.com
Tue May 15 08:43:14 PDT 2012


On 05/14/2012 10:23 PM, Peter Hutterer wrote:
> last.valuators contains the transformed valuators of the device. If the
> device submits events with x/y missing, we need to get that from
> last.valuators and undo the transformation to that axis.
> 
> X.Org Bug 49347 <http://bugs.freedesktop.org/show_bug.cgi?id=49347>
> 
> Signed-off-by: Peter Hutterer <peter.hutterer at who-t.net>
> ---
> Changes to v1:
> - exit early if neither x or y are set
> 
>  dix/getevents.c |   25 +++++++++++++++++++++----
>  1 file changed, 21 insertions(+), 4 deletions(-)
> 
> diff --git a/dix/getevents.c b/dix/getevents.c
> index c960d44..ae4112f 100644
> --- a/dix/getevents.c
> +++ b/dix/getevents.c
> @@ -1166,16 +1166,33 @@ static void
>  transformAbsolute(DeviceIntPtr dev, ValuatorMask *mask)
>  {
>      double x, y, ox, oy;
> +    int has_x, has_y;
> +
> +    has_x = valuator_mask_fetch_double(mask, 0, &ox);
> +    has_y = valuator_mask_fetch_double(mask, 1, &oy);
> +
> +    if (!has_x && !has_y)
> +        return;
> +
> +    if (!has_x || !has_y) {
> +        struct pixman_f_transform invert;
> +
> +        /* undo transformation from last event */
> +        ox = dev->last.valuators[0];
> +        oy = dev->last.valuators[1];
> +
> +        pixman_f_transform_invert(&invert, &dev->transform);
> +        transform(&invert, &ox, &oy);
> +
> +        x = ox;
> +        y = oy;
> +    }
>  
>      if (valuator_mask_isset(mask, 0))
>          ox = x = valuator_mask_get_double(mask, 0);
> -    else
> -        ox = x = dev->last.valuators[0];
>  
>      if (valuator_mask_isset(mask, 1))
>          oy = y = valuator_mask_get_double(mask, 1);
> -    else
> -        oy = y = dev->last.valuators[1];
>  
>      transform(&dev->transform, &x, &y);
>  

Looks good! Let's hope we don't hit any algorithmic instability with the
transformation inversion...

Reviewed-by: Chase Douglas <chase.douglas at canonical.com>


More information about the xorg-devel mailing list