hi Hans,
thank you for the patch. please use get_maintainer.pl to find all email
addresses to send this to. Sending this to linux-media only is
incomplete.
other than that, the patch lgtm and if you resend, you can add
Reviewed-by: Martin Kepplinger <martink@xxxxxxxxx>
thank you again,
martin
Am Dienstag, dem 18.04.2023 um 09:46 +0200 schrieb Hans Verkuil:
> Initialize hs_settle to 0 to avoid this compiler warning:
>
> imx8mq-mipi-csi2.c: In function
> 'imx8mq_mipi_csi_start_stream.part.0':
> imx8mq-mipi-csi2.c:91:55: warning: 'hs_settle' may be used
> uninitialized [-Wmaybe-uninitialized]
> 91 | #define GPR_CSI2_1_S_PRG_RXHS_SETTLE(x) (((x) & 0x3f) << 2)
> | ^~
> imx8mq-mipi-csi2.c:357:13: note: 'hs_settle' was declared here
> 357 | u32 hs_settle;
> | ^~~~~~~~~
>
> It's a false positive, but it is too complicated for the compiler to
> detect that.
>
> Signed-off-by: Hans Verkuil <hverkuil-cisco@xxxxxxxxx>
> ---
> diff --git a/drivers/staging/media/imx/imx8mq-mipi-csi2.c
> b/drivers/staging/media/imx/imx8mq-mipi-csi2.c
> index 32700cb8bc4d..ca2efcc21efe 100644
> --- a/drivers/staging/media/imx/imx8mq-mipi-csi2.c
> +++ b/drivers/staging/media/imx/imx8mq-mipi-csi2.c
> @@ -354,7 +354,7 @@ static int imx8mq_mipi_csi_start_stream(struct
> csi_state *state,
> struct v4l2_subdev_state
> *sd_state)
> {
> int ret;
> - u32 hs_settle;
> + u32 hs_settle = 0;
>
> ret = imx8mq_mipi_csi_sw_reset(state);
> if (ret)
