Re: [PATCH 1/1] Input: mma8450 - fix signed 12bits to 32bitsconversion

[Dmitry Torokhov]

Hi Seb,

On Wed, Mar 27, 2013 at 09:17:43AM +0100, seb wrote:
> Event value is wrong. Should be in range -2048 to 2047, but is in range 0 to 4095.
> Use int8_t to int conversion and remove 0xfff mask.
> 
> Signed-off-by: seb <sebastien.royen@xxxxxxxxxxxx>
> ---
>  drivers/input/misc/mma8450.c |   15 ++++++++++-----
>  1 file changed, 10 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/input/misc/mma8450.c b/drivers/input/misc/mma8450.c
> index 480557f..c3781a1 100644
> --- a/drivers/input/misc/mma8450.c
> +++ b/drivers/input/misc/mma8450.c
> @@ -110,7 +110,7 @@ static void mma8450_poll(struct input_polled_dev *dev)
>  	struct mma8450 *m = dev->private;
>  	int x, y, z;
>  	int ret;
> -	u8 buf[6];
> +	int8_t buf[6];
>  
>  	ret = mma8450_read(m, MMA8450_STATUS);
>  	if (ret < 0)
> @@ -119,13 +119,18 @@ static void mma8450_poll(struct input_polled_dev *dev)
>  	if (!(ret & MMA8450_STATUS_ZXYDR))
>  		return;
>  
> -	ret = mma8450_read_block(m, MMA8450_OUT_X_LSB, buf, sizeof(buf));
> +	ret = mma8450_read_block(m, MMA8450_OUT_X_LSB, (u8*)buf, sizeof(buf));
>  	if (ret < 0)
>  		return;
>  
> -	x = ((buf[1] << 4) & 0xff0) | (buf[0] & 0xf);
> -	y = ((buf[3] << 4) & 0xff0) | (buf[2] & 0xf);
> -	z = ((buf[5] << 4) & 0xff0) | (buf[4] & 0xf);
> +	/* convert 8 MSB from int8_t to int */
> +	x = buf[1];
> +	y = buf[3];
> +	z = buf[5];
> +	/* add 4 LSB */
> +	x = (x << 4) | (buf[0] & 0xf);
> +	y = (y << 4) | (buf[2] & 0xf);
> +	z = (z << 4) | (buf[4] & 0xf);

Should we just say:

	x = ((int)(s8)buf[1] << 4) | (buf[0] & 0xf);
	y = ...
	z = ...

and leave the rest as is?

Thanks.

-- 
Dmitry
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/