Hi Martin, On Wed, Jul 28, 2021 at 10:54:23AM +0200, Martin Kepplinger wrote: > Am Sonntag, dem 16.05.2021 um 05:42 +0300 schrieb Laurent Pinchart: > > Sample code from NXP, as well as experiments on i.MX8MM with RAW10 > > capture with an OV5640 sensor connected over CSI-2, showed that the > > TWO_8BIT_SENSOR field of the CSICR3 register needs to be set for > > formats > > larger than 8 bits. Do so, even if the reference manual doesn't > > clearly > > describe the effect of the field. > > > > Signed-off-by: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx> > > --- > > drivers/staging/media/imx/imx7-media-csi.c | 6 +++++- > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/staging/media/imx/imx7-media-csi.c > > b/drivers/staging/media/imx/imx7-media-csi.c > > index f85a2f5f1413..256b9aa978f0 100644 > > --- a/drivers/staging/media/imx/imx7-media-csi.c > > +++ b/drivers/staging/media/imx/imx7-media-csi.c > > @@ -422,6 +422,7 @@ static void imx7_csi_configure(struct imx7_csi > > *csi) > > int width = out_pix->width; > > u32 stride = 0; > > u32 cr1, cr18; > > + u32 cr3 = 0; > > > > cr18 = imx7_csi_reg_read(csi, CSI_CSICR18); > > > > @@ -464,6 +465,7 @@ static void imx7_csi_configure(struct imx7_csi > > *csi) > > case MEDIA_BUS_FMT_SGBRG10_1X10: > > case MEDIA_BUS_FMT_SGRBG10_1X10: > > case MEDIA_BUS_FMT_SRGGB10_1X10: > > + cr3 |= BIT_TWO_8BIT_SENSOR; > > cr18 |= BIT_MIPI_DATA_FORMAT_RAW10; > > break; > > case MEDIA_BUS_FMT_Y12_1X12: > > @@ -471,6 +473,7 @@ static void imx7_csi_configure(struct imx7_csi > > *csi) > > case MEDIA_BUS_FMT_SGBRG12_1X12: > > case MEDIA_BUS_FMT_SGRBG12_1X12: > > case MEDIA_BUS_FMT_SRGGB12_1X12: > > + cr3 |= BIT_TWO_8BIT_SENSOR; > > cr18 |= BIT_MIPI_DATA_FORMAT_RAW12; > > break; > > case MEDIA_BUS_FMT_Y14_1X14: > > @@ -478,6 +481,7 @@ static void imx7_csi_configure(struct imx7_csi > > *csi) > > case MEDIA_BUS_FMT_SGBRG14_1X14: > > case MEDIA_BUS_FMT_SGRBG14_1X14: > > case MEDIA_BUS_FMT_SRGGB14_1X14: > > + cr3 |= BIT_TWO_8BIT_SENSOR; > > cr18 |= BIT_MIPI_DATA_FORMAT_RAW14; > > break; > > /* > > @@ -510,7 +514,7 @@ static void imx7_csi_configure(struct imx7_csi > > *csi) > > > > imx7_csi_reg_write(csi, cr1, CSI_CSICR1); > > imx7_csi_reg_write(csi, BIT_DMA_BURST_TYPE_RFF_INCR16, > > CSI_CSICR2); > > - imx7_csi_reg_write(csi, BIT_FRMCNT_RST, CSI_CSICR3); > > + imx7_csi_reg_write(csi, BIT_FRMCNT_RST | cr3, CSI_CSICR3); > > imx7_csi_reg_write(csi, cr18, CSI_CSICR18); > > > > imx7_csi_reg_write(csi, (width * out_pix->height) >> 2, > > CSI_CSIRXCNT); > > this patch series looks good to me - I need it for the driver to run on > imx8mq. > > Reviewed-by: Martin Kepplinger <martin.kepplinger@xxxxxxx> > > Could you either rebase and resend as non-RFC or queue them up more > directly? I've just sent a pull request that includes v1.1 of this patch. I'm sorry I missed your Reviewed-by tag :-S https://lore.kernel.org/linux-media/YQFY5FS8v3Y3KkEA@xxxxxxxxxxxxxxxxxxxxxxxxxx/T/#u -- Regards, Laurent Pinchart
