gpu: ipu-v3: Add more planar formats support

Adds support for the following planar and partial-planar formats:

YUV422
NV12
NV16

Signed-off-by: Dmitry Eremin-Solenikov <dmitry_eremin@mentor.com>
Signed-off-by: Mohsin Kazmi <mohsin_kazmi@mentor.com>
Signed-off-by: Steve Longerbeam <steve_longerbeam@mentor.com>

Unified base offset and Y plane offset into a single variable,
moved all ipu_cpmem_set_buffer calls to a single location.
Removed NV21 and NV61 for now. The IDMAC doesn't understand U/V
order for chroma interleaved formats, so we'd need to work around
this by implenting U/V switching via the CSC unit.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

2 files changed, 120 insertions, 18 deletions

diff --git a/drivers/gpu/ipu-v3/ipu-common.c b/drivers/gpu/ipu-v3/ipu-common.c
index 28be741..5c3d526 100644
--- a/drivers/gpu/ipu-v3/ipu-common.c
+++ b/drivers/gpu/ipu-v3/ipu-common.c
@@ -74,6 +74,12 @@ enum ipu_color_space ipu_drm_fourcc_to_colorspace(u32 drm_fourcc)
 	case DRM_FORMAT_UYVY:
 	case DRM_FORMAT_YUV420:
 	case DRM_FORMAT_YVU420:
+	case DRM_FORMAT_YUV422:
+	case DRM_FORMAT_YVU422:
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_NV21:
+	case DRM_FORMAT_NV16:
+	case DRM_FORMAT_NV61:
 		return IPUV3_COLORSPACE_YUV;
 	default:
 		return IPUV3_COLORSPACE_UNKNOWN;
@@ -86,8 +92,13 @@ enum ipu_color_space ipu_pixelformat_to_colorspace(u32 pixelformat)
 	switch (pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
 	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
 	case V4L2_PIX_FMT_UYVY:
 	case V4L2_PIX_FMT_YUYV:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
 		return IPUV3_COLORSPACE_YUV;
 	case V4L2_PIX_FMT_RGB32:
 	case V4L2_PIX_FMT_BGR32:
@@ -106,6 +117,11 @@ bool ipu_pixelformat_is_planar(u32 pixelformat)
 	switch (pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
 	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
 		return true;
 	}
 
@@ -131,6 +147,11 @@ int ipu_stride_to_bytes(u32 pixel_stride, u32 pixelformat)
 	switch (pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
 	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV21:
+	case V4L2_PIX_FMT_NV16:
+	case V4L2_PIX_FMT_NV61:
 		/*
 		 * for the planar YUV formats, the stride passed to
 		 * cpmem must be the stride in bytes of the Y plane.
diff --git a/drivers/gpu/ipu-v3/ipu-cpmem.c b/drivers/gpu/ipu-v3/ipu-cpmem.c
index cfe2f53..c55f362 100644
--- a/drivers/gpu/ipu-v3/ipu-cpmem.c
+++ b/drivers/gpu/ipu-v3/ipu-cpmem.c
@@ -193,8 +193,14 @@ static int v4l2_pix_fmt_to_drm_fourcc(u32 pixelformat)
 		return DRM_FORMAT_YUYV;
 	case V4L2_PIX_FMT_YUV420:
 		return DRM_FORMAT_YUV420;
+	case V4L2_PIX_FMT_YUV422P:
+		return DRM_FORMAT_YUV422;
 	case V4L2_PIX_FMT_YVU420:
 		return DRM_FORMAT_YVU420;
+	case V4L2_PIX_FMT_NV12:
+		return DRM_FORMAT_NV12;
+	case V4L2_PIX_FMT_NV16:
+		return DRM_FORMAT_NV16;
 	}
 
 	return -EINVAL;
@@ -394,6 +400,7 @@ void ipu_cpmem_set_yuv_planar_full(struct ipuv3_channel *ch,
 {
 	switch (pixel_format) {
 	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YUV422P:
 		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, (stride / 2) - 1);
 		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, u_offset / 8);
 		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, v_offset / 8);
@@ -403,6 +410,12 @@ void ipu_cpmem_set_yuv_planar_full(struct ipuv3_channel *ch,
 		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, v_offset / 8);
 		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, u_offset / 8);
 		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV16:
+		ipu_ch_param_write_field(ch, IPU_FIELD_SLUV, stride - 1);
+		ipu_ch_param_write_field(ch, IPU_FIELD_UBO, u_offset / 8);
+		ipu_ch_param_write_field(ch, IPU_FIELD_VBO, u_offset / 8);
+		break;
 	}
 }
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar_full);
@@ -422,6 +435,19 @@ void ipu_cpmem_set_yuv_planar(struct ipuv3_channel *ch,
 		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
 					      u_offset, v_offset);
 		break;
+	case V4L2_PIX_FMT_YUV422P:
+		uv_stride = stride / 2;
+		u_offset = stride * height;
+		v_offset = u_offset + (uv_stride * height);
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_NV16:
+		u_offset = stride * height;
+		ipu_cpmem_set_yuv_planar_full(ch, pixel_format, stride,
+					      u_offset, 0);
+		break;
 	}
 }
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_yuv_planar);
@@ -475,11 +501,20 @@ static const struct ipu_rgb def_bgr_16 = {
 };
 
 #define Y_OFFSET(pix, x, y)	((x) + pix->width * (y))
-#define U_OFFSET(pix, x, y)	((pix->width * pix->height) + \
-					(pix->width * (y) / 4) + (x) / 2)
-#define V_OFFSET(pix, x, y)	((pix->width * pix->height) + \
-					(pix->width * pix->height / 4) + \
-					(pix->width * (y) / 4) + (x) / 2)
+#define U_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * (y) / 4) + (x) / 2)
+#define V_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * pix->height / 4) +	\
+				 (pix->width * (y) / 4) + (x) / 2)
+#define U2_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * (y) / 2) + (x) / 2)
+#define V2_OFFSET(pix, x, y)	((pix->width * pix->height) +		\
+				 (pix->width * pix->height / 2) +	\
+				 (pix->width * (y) / 2) + (x) / 2)
+#define UV_OFFSET(pix, x, y)	((pix->width * pix->height) +	\
+				 (pix->width * (y) / 2) + (x))
+#define UV2_OFFSET(pix, x, y)	((pix->width * pix->height) +	\
+				 (pix->width * y) + (x))
 
 int ipu_cpmem_set_fmt(struct ipuv3_channel *ch, u32 drm_fourcc)
 {
@@ -491,6 +526,25 @@ int ipu_cpmem_set_fmt(struct ipuv3_channel *ch, u32 drm_fourcc)
 		/* burst size */
 		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
 		break;
+	case DRM_FORMAT_YUV422:
+	case DRM_FORMAT_YVU422:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 1);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_NV12:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 4);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
+	case DRM_FORMAT_NV16:
+		/* pix format */
+		ipu_ch_param_write_field(ch, IPU_FIELD_PFS, 3);
+		/* burst size */
+		ipu_ch_param_write_field(ch, IPU_FIELD_NPB, 31);
+		break;
 	case DRM_FORMAT_UYVY:
 		/* bits/pixel */
 		ipu_ch_param_write_field(ch, IPU_FIELD_BPP, 3);
@@ -538,7 +592,7 @@ EXPORT_SYMBOL_GPL(ipu_cpmem_set_fmt);
 int ipu_cpmem_set_image(struct ipuv3_channel *ch, struct ipu_image *image)
 {
 	struct v4l2_pix_format *pix = &image->pix;
-	int offset, y_offset, u_offset, v_offset;
+	int offset, u_offset, v_offset;
 
 	pr_debug("%s: resolution: %dx%d stride: %d\n",
 		 __func__, pix->width, pix->height,
@@ -552,43 +606,70 @@ int ipu_cpmem_set_image(struct ipuv3_channel *ch, struct ipu_image *image)
 	switch (pix->pixelformat) {
 	case V4L2_PIX_FMT_YUV420:
 	case V4L2_PIX_FMT_YVU420:
-		y_offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
 		u_offset = U_OFFSET(pix, image->rect.left,
-				    image->rect.top) - y_offset;
+				    image->rect.top) - offset;
 		v_offset = V_OFFSET(pix, image->rect.left,
-				    image->rect.top) - y_offset;
+				    image->rect.top) - offset;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_YUV422P:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = U2_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+		v_offset = V2_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV12:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = UV_OFFSET(pix, image->rect.left,
+				     image->rect.top) - offset;
+		v_offset = 0;
 
 		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
-				pix->bytesperline, u_offset, v_offset);
-		ipu_cpmem_set_buffer(ch, 0, image->phys0 + y_offset);
-		ipu_cpmem_set_buffer(ch, 1, image->phys1 + y_offset);
+					      pix->bytesperline,
+					      u_offset, v_offset);
+		break;
+	case V4L2_PIX_FMT_NV16:
+		offset = Y_OFFSET(pix, image->rect.left, image->rect.top);
+		u_offset = UV2_OFFSET(pix, image->rect.left,
+				      image->rect.top) - offset;
+		v_offset = 0;
+
+		ipu_cpmem_set_yuv_planar_full(ch, pix->pixelformat,
+					      pix->bytesperline,
+					      u_offset, v_offset);
 		break;
 	case V4L2_PIX_FMT_UYVY:
 	case V4L2_PIX_FMT_YUYV:
 	case V4L2_PIX_FMT_RGB565:
 		offset = image->rect.left * 2 +
 			image->rect.top * pix->bytesperline;
-		ipu_cpmem_set_buffer(ch, 0, image->phys0 + offset);
-		ipu_cpmem_set_buffer(ch, 1, image->phys1 + offset);
 		break;
 	case V4L2_PIX_FMT_RGB32:
 	case V4L2_PIX_FMT_BGR32:
 		offset = image->rect.left * 4 +
 			image->rect.top * pix->bytesperline;
-		ipu_cpmem_set_buffer(ch, 0, image->phys0 + offset);
-		ipu_cpmem_set_buffer(ch, 1, image->phys1 + offset);
 		break;
 	case V4L2_PIX_FMT_RGB24:
 	case V4L2_PIX_FMT_BGR24:
 		offset = image->rect.left * 3 +
 			image->rect.top * pix->bytesperline;
-		ipu_cpmem_set_buffer(ch, 0, image->phys0 + offset);
-		ipu_cpmem_set_buffer(ch, 1, image->phys1 + offset);
 		break;
 	default:
 		return -EINVAL;
 	}
 
+	ipu_cpmem_set_buffer(ch, 0, image->phys0 + offset);
+	ipu_cpmem_set_buffer(ch, 1, image->phys1 + offset);
+
 	return 0;
 }
 EXPORT_SYMBOL_GPL(ipu_cpmem_set_image);