b2r2:Migrate driver to kernel 3.0

Migrate the following b2r2 driver patches to kernel3.0
dbc424d b2r2: Add optimized path for VUY(A)888(8) 24/32-bit formats
2142b2d b2r2: Add VUY(A)888(8) 24/32-bit formats
797c68c [b2r2] Fix QA tools warning in b2r2_node_split.c

ST-Ericsson ID: 352334
ST-Ericsson Linux next: NA
ST-Ericsson FOSS-OUT ID: NA

Change-Id: I5f3160a9568ad17eb05eae0b7b8adba146153ea9
Signed-off-by: Naveen Kumar Gaddipati <naveen.gaddipati@stericsson.com>
Reviewed-on: http://gerrit.lud.stericsson.com/gerrit/29563

7 files changed, 341 insertions, 88 deletions

diff --git a/drivers/video/b2r2/b2r2_blt_main.c b/drivers/video/b2r2/b2r2_blt_main.c
index a2b21f4a2c1..097d2854a83 100644
--- a/drivers/video/b2r2/b2r2_blt_main.c
+++ b/drivers/video/b2r2/b2r2_blt_main.c
@@ -339,7 +339,7 @@ static int b2r2_blt_release(struct inode *inode, struct file *filp)
 		job = b2r2_core_job_find_first_with_tag((int) instance);
 		while (job) {
 			b2r2_core_job_cancel(job);
-			/* This release matches addref in b2r2_core_job_find... */
+			/* Matches addref in b2r2_core_job_find... */
 			b2r2_core_job_release(job, __func__);
 			job = b2r2_core_job_find_first_with_tag((int) instance);
 		}
@@ -483,8 +483,8 @@ static long b2r2_blt_ioctl(struct file *file,
 			b2r2_log_info("b2r2_blt=%d Going generic.\n", ret);
 			request_gen = kmalloc(sizeof(*request_gen), GFP_KERNEL);
 			if (!request_gen) {
-				b2r2_log_err(
-					"%s: Failed to alloc mem for request_gen\n", __func__);
+				b2r2_log_err("%s: Failed to alloc mem for "
+					"request_gen\n", __func__);
 				return -ENOMEM;
 			}
 
@@ -513,19 +513,27 @@ static long b2r2_blt_ioctl(struct file *file,
 			 * make a copy that the HW can use.
 			 */
 			if ((request_gen->user_req.flags &
-					B2R2_BLT_FLAG_CLUT_COLOR_CORRECTION) != 0) {
-				request_gen->clut = dma_alloc_coherent(b2r2_blt_device(),
-					CLUT_SIZE, &(request_gen->clut_phys_addr),
+					B2R2_BLT_FLAG_CLUT_COLOR_CORRECTION)
+					!= 0) {
+				request_gen->clut =
+					dma_alloc_coherent(b2r2_blt_device(),
+					CLUT_SIZE,
+					&(request_gen->clut_phys_addr),
 					GFP_DMA | GFP_KERNEL);
 				if (request_gen->clut == NULL) {
-					b2r2_log_err("%s CLUT allocation failed.\n", __func__);
+					b2r2_log_err("%s CLUT allocation "
+						"failed.\n", __func__);
 					kfree(request_gen);
 					return -ENOMEM;
 				}
 
-				if (copy_from_user(request_gen->clut, request_gen->user_req.clut, CLUT_SIZE)) {
-					b2r2_log_err("%s: CLUT copy_from_user failed\n", __func__);
-					dma_free_coherent(b2r2_blt_device(), CLUT_SIZE, request_gen->clut,
+				if (copy_from_user(request_gen->clut,
+						request_gen->user_req.clut,
+						CLUT_SIZE)) {
+					b2r2_log_err("%s: CLUT copy_from_user "
+						"failed\n", __func__);
+					dma_free_coherent(b2r2_blt_device(),
+						CLUT_SIZE, request_gen->clut,
 						request_gen->clut_phys_addr);
 					request_gen->clut = NULL;
 					request_gen->clut_phys_addr = 0;
@@ -537,7 +545,8 @@ static long b2r2_blt_ioctl(struct file *file,
 			request_gen->profile = is_profiler_registered_approx();
 
 			ret = b2r2_generic_blt(instance, request_gen);
-			b2r2_log_info("\nb2r2_generic_blt=%d Generic done.\n", ret);
+			b2r2_log_info("\nb2r2_generic_blt=%d Generic done.\n",
+				ret);
 		}
 #endif /* CONFIG_B2R2_GENERIC_FALLBACK */
 
@@ -1339,9 +1348,8 @@ static void tile_job_callback_gen(struct b2r2_core_job *job)
 
 #ifdef CONFIG_DEBUG_FS
 	/* Notify if a tile job is cancelled */
-	if (job->job_state == B2R2_CORE_JOB_CANCELED) {
+	if (job->job_state == B2R2_CORE_JOB_CANCELED)
 		b2r2_log_info("%s: Tile job cancelled:\n", __func__);
-	}
 #endif
 
 	/* Local addref / release within this func */
@@ -1832,19 +1840,24 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 			x = -dst_rect->x;
 
 		for (; x < dst_rect->width &&
-				x + dst_rect->x < dst_img_width; x += tmp_buf_width) {
+				x + dst_rect->x < dst_img_width;
+				x += tmp_buf_width) {
 			/*
 			 * Tile jobs are freed by the supplied release function
 			 * when ref_count on a tile_job reaches zero.
 			 */
-			struct b2r2_core_job *tile_job = kmalloc(sizeof(*tile_job), GFP_KERNEL);
+			struct b2r2_core_job *tile_job =
+				kmalloc(sizeof(*tile_job), GFP_KERNEL);
 			if (tile_job == NULL) {
 				/*
-				 * Skip this tile. Do not abort, just hope for better luck
-				 * with rest of the tiles. Memory might become available.
+				 * Skip this tile. Do not abort,
+				 * just hope for better luck
+				 * with rest of the tiles.
+				 * Memory might become available.
 				 */
 				b2r2_log_info("%s: Failed to alloc job. "
-					"Skipping tile at (x, y)=(%d, %d)\n", __func__, x, y);
+					"Skipping tile at (x, y)=(%d, %d)\n",
+					__func__, x, y);
 				continue;
 			}
 			tile_job->tag = request->job.tag;
@@ -1865,7 +1878,8 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 				 * Only a part of the tile can be written.
 				 * Limit imposed by buffer size.
 				 */
-				dst_rect_tile.width = dst_img_width - (x + dst_rect->x);
+				dst_rect_tile.width =
+					dst_img_width - (x + dst_rect->x);
 			} else if (x + tmp_buf_width > dst_rect->width) {
 				/*
 				 * Only a part of the tile can be written.
@@ -1877,10 +1891,12 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 				dst_rect_tile.width = tmp_buf_width;
 			}
 			/*
-			 * Where applicable, calculate area in src buffer that is needed
-			 * to generate the specified part of destination rectangle.
+			 * Where applicable, calculate area in src buffer
+			 * that is needed to generate the specified part
+			 * of destination rectangle.
 			 */
-			b2r2_generic_set_areas(request, request->first_node, &dst_rect_tile);
+			b2r2_generic_set_areas(request,
+				request->first_node, &dst_rect_tile);
 			/* Submit the job */
 			b2r2_log_info("%s: Submitting job\n", __func__);
 
@@ -1940,7 +1956,8 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 			x + dst_rect->x < dst_img_width; x += tmp_buf_width) {
 		struct b2r2_core_job *tile_job = NULL;
 		if (x + tmp_buf_width < dst_rect->width &&
-				x + dst_rect->x + tmp_buf_width < dst_img_width) {
+				x + dst_rect->x + tmp_buf_width <
+				dst_img_width) {
 			/*
 			 * Tile jobs are freed by the supplied release function
 			 * when ref_count on a tile_job reaches zero.
@@ -1987,26 +2004,29 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 		}
 		/*
 		 * y is now the last row. Either because the whole dst_rect
-		 * has been processed, or because the last row that will be written
-		 * to dst_img has been reached. Limits imposed in the same way
-		 * as for width.
+		 * has been processed, or because the last row that will be
+		 * written to dst_img has been reached. Limits imposed in
+		 * the same way as for width.
 		 */
 		dst_rect_tile.y = y;
 		if (y + dst_rect->y + tmp_buf_height > dst_img_height)
-			dst_rect_tile.height = dst_img_height - (y + dst_rect->y);
+			dst_rect_tile.height =
+				dst_img_height - (y + dst_rect->y);
 		else if (y + tmp_buf_height > dst_rect->height)
 			dst_rect_tile.height = dst_rect->height - y;
 		else
 			dst_rect_tile.height = tmp_buf_height;
 
-		b2r2_generic_set_areas(request, request->first_node, &dst_rect_tile);
+		b2r2_generic_set_areas(request,
+			request->first_node, &dst_rect_tile);
 
 		b2r2_log_info("%s: Submitting job\n", __func__);
 		inc_stat(&stat_n_in_blt_add);
 
 		mutex_lock(&instance->lock);
 		if (x + tmp_buf_width < dst_rect->width &&
-				x + dst_rect->x + tmp_buf_width < dst_img_width) {
+				x + dst_rect->x + tmp_buf_width <
+				dst_img_width) {
 			request_id = b2r2_core_job_add(tile_job);
 		} else {
 			/*
@@ -2036,7 +2056,8 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 
 		inc_stat(&stat_n_in_blt_wait);
 		if (x + tmp_buf_width < dst_rect->width &&
-				x + dst_rect->x + tmp_buf_width < dst_img_width) {
+				x + dst_rect->x + tmp_buf_width <
+				dst_img_width) {
 			ret = b2r2_core_job_wait(tile_job);
 		} else {
 			/*
@@ -2056,8 +2077,10 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 				"%s: Synchronous wait done\n", __func__);
 
 			if (x + tmp_buf_width < dst_rect->width &&
-					x + dst_rect->x + tmp_buf_width < dst_img_width)
-				nsec_active_in_b2r2 += tile_job->nsec_active_in_hw;
+					x + dst_rect->x + tmp_buf_width <
+					dst_img_width)
+				nsec_active_in_b2r2 +=
+					tile_job->nsec_active_in_hw;
 			else
 				nsec_active_in_b2r2 +=
 					request->job.nsec_active_in_hw;
@@ -2069,7 +2092,8 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 		 * when the last tile is processed.
 		 */
 		if (x + tmp_buf_width < dst_rect->width &&
-				x + dst_rect->x + tmp_buf_width < dst_img_width) {
+				x + dst_rect->x + tmp_buf_width <
+				dst_img_width) {
 			b2r2_core_job_release(tile_job, __func__);
 		} else {
 			b2r2_core_job_release(&request->job, __func__);
@@ -2100,7 +2124,8 @@ static int b2r2_generic_blt(struct b2r2_blt_instance *instance,
 
 	if (request->profile) {
 		request->nsec_active_in_cpu =
-			(s32)((u32)task_sched_runtime(current) - thread_runtime_at_start);
+			(s32)((u32)task_sched_runtime(current) -
+			thread_runtime_at_start);
 		request->total_time_nsec =
 			(s32)(b2r2_get_curr_nsec() - request->start_time_nsec);
 		request->job.nsec_active_in_hw = nsec_active_in_b2r2;
@@ -2433,7 +2458,8 @@ static int resolve_buf(struct b2r2_blt_img *img,
 			if (file == NULL)
 				return -EINVAL;
 #ifdef CONFIG_FB
-			if (MAJOR(file->f_dentry->d_inode->i_rdev) == FB_MAJOR) {
+			if (MAJOR(file->f_dentry->d_inode->i_rdev) ==
+					FB_MAJOR) {
 				/*
 				 * This is a frame buffer device, find fb_info
 				 * (OK to do it like this, no locking???)
@@ -2503,7 +2529,8 @@ static int resolve_buf(struct b2r2_blt_img *img,
  * @resolved_buf: Gathered info (physical address etc.) about buffer
  * @is_dst: true if the buffer is a destination buffer, false if the buffer is a
  *          source buffer.
- * @rect: rectangle in the image buffer that should be synced. NULL the buffer is a source mask.
+ * @rect: rectangle in the image buffer that should be synced.
+ *        NULL if the buffer is a source mask.
  * @img_width: width of the complete image buffer
  * @fmt: buffer format
 */
@@ -2554,11 +2581,15 @@ static void sync_buf(struct b2r2_blt_img *img,
 			(img->fmt ==
 				B2R2_BLT_FMT_YUV422_PACKED_SEMIPLANAR_MB_STE)) {
 		sa.start = (unsigned long)resolved->virtual_address;
-		sa.end = (unsigned long)resolved->virtual_address + img->buf.len;
+		sa.end = (unsigned long)resolved->virtual_address +
+			img->buf.len;
 		start_phys = resolved->physical_address;
 		end_phys = resolved->physical_address + img->buf.len;
 	} else {
-		/* buffer is not a src_mask so make use of rect when clean & flush caches*/
+		/*
+		 * buffer is not a src_mask so make use of rect when
+		 * clean & flush caches
+		 */
 		u32 bpp;	/* Bits per pixel */
 		u32 pitch;
 
@@ -2573,11 +2604,13 @@ static void sync_buf(struct b2r2_blt_img *img,
 		case B2R2_BLT_FMT_24_BIT_RGB888:   /* Fall through */
 		case B2R2_BLT_FMT_24_BIT_ARGB8565: /* Fall through */
 		case B2R2_BLT_FMT_24_BIT_YUV888:
+		case B2R2_BLT_FMT_24_BIT_VUY888:
 			bpp =  24;
 			break;
 		case B2R2_BLT_FMT_32_BIT_ARGB8888: /* Fall through */
 		case B2R2_BLT_FMT_32_BIT_ABGR8888: /* Fall through */
 		case B2R2_BLT_FMT_32_BIT_AYUV8888:
+		case B2R2_BLT_FMT_32_BIT_VUYA8888:
 			bpp = 32;
 			break;
 		default:
@@ -2930,7 +2963,8 @@ static int debugfs_b2r2_blt_request_read(struct file *filp, char __user *buf,
 		goto out;
 	}
 
-	dev_size = sprintf_req(&debugfs_latest_request, Buf, sizeof(char) * 4096);
+	dev_size = sprintf_req(&debugfs_latest_request, Buf,
+		sizeof(char) * 4096);
 
 	/* No more to read if offset != 0 */
 	if (*f_pos > dev_size)
diff --git a/drivers/video/b2r2/b2r2_generic.c b/drivers/video/b2r2/b2r2_generic.c
index 738e11bf4ba..5941e39be91 100644
--- a/drivers/video/b2r2/b2r2_generic.c
+++ b/drivers/video/b2r2/b2r2_generic.c
@@ -180,11 +180,13 @@ static inline enum b2r2_native_fmt to_native_fmt(enum b2r2_blt_fmt fmt)
 		return B2R2_NATIVE_ARGB8565;
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 		return B2R2_NATIVE_RGB888;
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 		return B2R2_NATIVE_YCBCR888;
 	case B2R2_BLT_FMT_32_BIT_ABGR8888: /* Not actually supported by HW */
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 		return B2R2_NATIVE_ARGB8888;
+	case B2R2_BLT_FMT_32_BIT_VUYA8888: /* fall through */
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
 		return B2R2_NATIVE_AYCBCR8888;
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
@@ -220,6 +222,7 @@ static inline enum b2r2_ty get_alpha_range(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_8_BIT_A8:
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 		return B2R2_TY_ALPHA_RANGE_255; /* 0 - 255 */
@@ -250,15 +253,15 @@ static unsigned int get_pitch(enum b2r2_blt_fmt format, u32 width)
 	case B2R2_BLT_FMT_16_BIT_ARGB4444:
 		return width * 2;
 		break;
-	case B2R2_BLT_FMT_24_BIT_RGB888: /* all 24 bits/pixel RGB formats */
+	case B2R2_BLT_FMT_24_BIT_RGB888: /* all 24 bits/pixel raster formats */
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 		return width * 3;
 		break;
-	case B2R2_BLT_FMT_32_BIT_ARGB8888: /* all 32 bits/pixel RGB formats */
+	case B2R2_BLT_FMT_32_BIT_ARGB8888: /* all 32 bits/pixel formats */
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
-		return width * 4;
-		break;
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
 		return width * 4;
 		break;
@@ -532,6 +535,8 @@ static void setup_fill_input_stage(const struct b2r2_blt_request *req,
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_YUV420_PACKED_PLANAR:
 	case B2R2_BLT_FMT_YUV422_PACKED_PLANAR:
 	case B2R2_BLT_FMT_YVU420_PACKED_PLANAR:
@@ -607,6 +612,8 @@ static void setup_fill_input_stage(const struct b2r2_blt_request *req,
 			switch (dst_img->fmt) {
 			case B2R2_BLT_FMT_24_BIT_YUV888:
 			case B2R2_BLT_FMT_32_BIT_AYUV8888:
+			case B2R2_BLT_FMT_24_BIT_VUY888:
+			case B2R2_BLT_FMT_32_BIT_VUYA8888:
 				node->node.GROUP0.B2R2_INS |=
 					B2R2_INS_IVMX_ENABLED;
 				node->node.GROUP0.B2R2_CIC |= B2R2_CIC_IVMX;
@@ -618,6 +625,27 @@ static void setup_fill_input_stage(const struct b2r2_blt_request *req,
 					B2R2_VMX2_BLT_YUV888_TO_RGB_601_VIDEO;
 				node->node.GROUP15.B2R2_VMX3 =
 					B2R2_VMX3_BLT_YUV888_TO_RGB_601_VIDEO;
+				/*
+				 * Re-arrange the color components from
+				 * VUY(A) to (A)YUV
+				 */
+				if (dst_img->fmt ==
+					B2R2_BLT_FMT_24_BIT_VUY888) {
+					u32 Y = src_color & 0xff;
+					u32 U = src_color & 0xff00;
+					u32 V = src_color & 0xff0000;
+					src_color = (Y << 16) | U | (V >> 16);
+				} else if (dst_img->fmt ==
+						B2R2_BLT_FMT_32_BIT_VUYA8888) {
+					u32 A = src_color & 0xff;
+					u32 Y = src_color & 0xff00;
+					u32 U = src_color & 0xff0000;
+					u32 V = src_color & 0xff000000;
+					src_color = (A << 24) |
+							(Y << 8) |
+							(U >> 8) |
+							(V >> 24);
+				}
 				break;
 			case B2R2_BLT_FMT_Y_CB_Y_CR:
 				/*
@@ -703,6 +731,7 @@ static void setup_input_stage(const struct b2r2_blt_request *req,
 	/* horizontal and vertical scan order for out_buf */
 	enum b2r2_ty dst_hso = B2R2_TY_HSO_LEFT_TO_RIGHT;
 	enum b2r2_ty dst_vso = B2R2_TY_VSO_TOP_TO_BOTTOM;
+	u32 endianness = 0;
 	u32 fctl = 0;
 	u32 rsf = 0;
 	u32 rzi = 0;
@@ -838,9 +867,13 @@ static void setup_input_stage(const struct b2r2_blt_request *req,
 		break;
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		/*
 		 * Set up IVMX.
-		 * Color components are laid out in memory as V, U, Y, (A)
+		 * For B2R2_BLT_FMT_32_BIT_YUV888 and
+		 * B2R2_BLT_FMT_32_BIT_AYUV8888
+		 * the color components are laid out in memory as V, U, Y, (A)
 		 * with V at the first byte (due to little endian addressing).
 		 * B2R2 expects them to be as U, Y, V, (A)
 		 * with U at the first byte.
@@ -855,6 +888,14 @@ static void setup_input_stage(const struct b2r2_blt_request *req,
 			B2R2_VMX2_BLT_YUV888_TO_RGB_601_VIDEO;
 		node->node.GROUP15.B2R2_VMX3 =
 			B2R2_VMX3_BLT_YUV888_TO_RGB_601_VIDEO;
+
+		/*
+		 * Re-arrange color components from VUY(A) to (A)YUV
+		 * for input VMX to work on them further.
+		 */
+		if (src_img->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+				src_img->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+			endianness = B2R2_TY_ENDIAN_BIG_NOT_LITTLE;
 		break;
 	case B2R2_BLT_FMT_YUV420_PACKED_PLANAR:
 	case B2R2_BLT_FMT_YUV422_PACKED_PLANAR:
@@ -1161,7 +1202,8 @@ static void setup_input_stage(const struct b2r2_blt_request *req,
 			to_native_fmt(src_img->fmt) |
 			get_alpha_range(src_img->fmt) |
 			B2R2_TY_HSO_LEFT_TO_RIGHT |
-			B2R2_TY_VSO_TOP_TO_BOTTOM;
+			B2R2_TY_VSO_TOP_TO_BOTTOM |
+			endianness;
 
 		node->node.GROUP0.B2R2_INS |= B2R2_INS_SOURCE_2_FETCH_FROM_MEM;
 		node->node.GROUP0.B2R2_CIC |= B2R2_CIC_SOURCE_2;
@@ -1240,6 +1282,7 @@ static void setup_dst_read_stage(const struct b2r2_blt_request *req,
 	const struct b2r2_blt_img *dst_img = &(req->user_req.dst_img);
 	u32 fctl = 0;
 	u32 rsf = 0;
+	u32 endianness = 0;
 	bool yuv_semi_planar =
 		dst_img->fmt == B2R2_BLT_FMT_YUV420_PACKED_SEMI_PLANAR ||
 		dst_img->fmt == B2R2_BLT_FMT_YUV422_PACKED_SEMI_PLANAR ||
@@ -1278,36 +1321,59 @@ static void setup_dst_read_stage(const struct b2r2_blt_request *req,
 		node->node.GROUP15.B2R2_VMX3 = B2R2_VMX3_RGB_TO_BGR;
 		break;
 	case B2R2_BLT_FMT_Y_CB_Y_CR:
+		/* Set up IVMX */
+		node->node.GROUP0.B2R2_INS |= B2R2_INS_IVMX_ENABLED;
+		node->node.GROUP0.B2R2_CIC |= B2R2_CIC_IVMX;
+		/*
+		 * Setup input VMX to convert YVU to RGB 601 VIDEO
+		 * Chroma components are swapped
+		 * so it is YVU and not YUV.
+		 */
+		node->node.GROUP15.B2R2_VMX0 = B2R2_VMX0_YVU_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX1 = B2R2_VMX1_YVU_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX2 = B2R2_VMX2_YVU_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX3 = B2R2_VMX3_YVU_TO_RGB_601_VIDEO;
+		break;
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
+		/* Set up IVMX */
+		node->node.GROUP0.B2R2_INS |= B2R2_INS_IVMX_ENABLED;
+		node->node.GROUP0.B2R2_CIC |= B2R2_CIC_IVMX;
+		node->node.GROUP15.B2R2_VMX0 = B2R2_VMX0_YUV_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX1 = B2R2_VMX1_YUV_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX2 = B2R2_VMX2_YUV_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX3 = B2R2_VMX3_YUV_TO_RGB_601_VIDEO;
+		break;
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
-		/* Set up IVMX */
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
+		/*
+		 * Set up IVMX.
+		 * For B2R2_BLT_FMT_32_BIT_YUV888 and
+		 * B2R2_BLT_FMT_32_BIT_AYUV8888
+		 * the color components are laid out in memory as V, U, Y, (A)
+		 * with V at the first byte (due to little endian addressing).
+		 * B2R2 expects them to be as U, Y, V, (A)
+		 * with U at the first byte.
+		 */
 		node->node.GROUP0.B2R2_INS |= B2R2_INS_IVMX_ENABLED;
 		node->node.GROUP0.B2R2_CIC |= B2R2_CIC_IVMX;
-		if (dst_img->fmt == B2R2_BLT_FMT_Y_CB_Y_CR) {
-			/*
-			 * Setup input VMX to convert YVU to RGB 601 VIDEO
-			 * Chroma components are swapped
-			 * so it is YVU and not YUV.
-			 */
-			node->node.GROUP15.B2R2_VMX0 =
-				B2R2_VMX0_YVU_TO_RGB_601_VIDEO;
-			node->node.GROUP15.B2R2_VMX1 =
-				B2R2_VMX1_YVU_TO_RGB_601_VIDEO;
-			node->node.GROUP15.B2R2_VMX2 =
-				B2R2_VMX2_YVU_TO_RGB_601_VIDEO;
-			node->node.GROUP15.B2R2_VMX3 =
-				B2R2_VMX3_YVU_TO_RGB_601_VIDEO;
-		} else {
-			node->node.GROUP15.B2R2_VMX0 =
-				B2R2_VMX0_YUV_TO_RGB_601_VIDEO;
-			node->node.GROUP15.B2R2_VMX1 =
-				B2R2_VMX1_YUV_TO_RGB_601_VIDEO;
-			node->node.GROUP15.B2R2_VMX2 =
-				B2R2_VMX2_YUV_TO_RGB_601_VIDEO;
-			node->node.GROUP15.B2R2_VMX3 =
-				B2R2_VMX3_YUV_TO_RGB_601_VIDEO;
-		}
+		node->node.GROUP15.B2R2_VMX0 =
+			B2R2_VMX0_BLT_YUV888_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX1 =
+			B2R2_VMX1_BLT_YUV888_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX2 =
+			B2R2_VMX2_BLT_YUV888_TO_RGB_601_VIDEO;
+		node->node.GROUP15.B2R2_VMX3 =
+			B2R2_VMX3_BLT_YUV888_TO_RGB_601_VIDEO;
+
+		/*
+		 * Re-arrange color components from VUY(A) to (A)YUV
+		 * for input VMX to work on them further.
+		 */
+		if (dst_img->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+				dst_img->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+			endianness = B2R2_TY_ENDIAN_BIG_NOT_LITTLE;
 		break;
 	case B2R2_BLT_FMT_YUV420_PACKED_PLANAR:
 	case B2R2_BLT_FMT_YUV422_PACKED_PLANAR:
@@ -1536,7 +1602,8 @@ static void setup_dst_read_stage(const struct b2r2_blt_request *req,
 			to_native_fmt(dst_img->fmt) |
 			get_alpha_range(dst_img->fmt) |
 			B2R2_TY_HSO_LEFT_TO_RIGHT |
-			B2R2_TY_VSO_TOP_TO_BOTTOM;
+			B2R2_TY_VSO_TOP_TO_BOTTOM |
+			endianness;
 
 		node->node.GROUP0.B2R2_INS |=
 			B2R2_INS_SOURCE_2_FETCH_FROM_MEM;
@@ -1689,6 +1756,7 @@ static void setup_writeback_stage(const struct b2r2_blt_request *req,
 
 	u32 dst_dither = 0;
 	u32 dst_pitch = 0;
+	u32 endianness = 0;
 
 	b2r2_log_info("%s ENTRY\n", __func__);
 
@@ -2057,7 +2125,9 @@ static void setup_writeback_stage(const struct b2r2_blt_request *req,
 				B2R2_VMX3_RGB_TO_YVU_601_VIDEO;
 			break;
 		case B2R2_BLT_FMT_24_BIT_YUV888: /* fall through */
-		case B2R2_BLT_FMT_32_BIT_AYUV8888:
+		case B2R2_BLT_FMT_32_BIT_AYUV8888: /* fall through */
+		case B2R2_BLT_FMT_24_BIT_VUY888: /* fall through */
+		case B2R2_BLT_FMT_32_BIT_VUYA8888:
 			node->node.GROUP0.B2R2_INS |= B2R2_INS_OVMX_ENABLED;
 			node->node.GROUP0.B2R2_CIC |= B2R2_CIC_OVMX;
 			node->node.GROUP16.B2R2_VMX0 =
@@ -2068,6 +2138,14 @@ static void setup_writeback_stage(const struct b2r2_blt_request *req,
 				B2R2_VMX2_RGB_TO_BLT_YUV888_601_VIDEO;
 			node->node.GROUP16.B2R2_VMX3 =
 				B2R2_VMX3_RGB_TO_BLT_YUV888_601_VIDEO;
+
+			/*
+			 * Re-arrange color components from (A)YUV to VUY(A)
+			 * when bytes are stored in memory.
+			 */
+			if (dst_fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+					dst_fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+				endianness = B2R2_TY_ENDIAN_BIG_NOT_LITTLE;
 			break;
 		default:
 			break;
@@ -2080,7 +2158,8 @@ static void setup_writeback_stage(const struct b2r2_blt_request *req,
 			get_alpha_range(dst_img->fmt) |
 			B2R2_TY_HSO_LEFT_TO_RIGHT |
 			B2R2_TY_VSO_TOP_TO_BOTTOM |
-			dst_dither;
+			dst_dither |
+			endianness;
 
 		node->node.GROUP0.B2R2_ACK = B2R2_ACK_MODE_BYPASS_S2_S3;
 		node->node.GROUP0.B2R2_INS |=
diff --git a/drivers/video/b2r2/b2r2_hw.h b/drivers/video/b2r2/b2r2_hw.h
index 0dd5f758eb5..d492168913a 100644
--- a/drivers/video/b2r2/b2r2_hw.h
+++ b/drivers/video/b2r2/b2r2_hw.h
@@ -680,4 +680,28 @@ enum b2r2_plug_page_size {
 #define B2R2_VMX2_BLT_YUV888_TO_RGB_601_VIDEO 0x2006f000
 #define B2R2_VMX3_BLT_YUV888_TO_RGB_601_VIDEO 0x34f21322
 
+/* VMX register values for YUV to BLT_YUV888 conversion */
+#define B2R2_VMX0_YUV_TO_BLT_YUV888 0x00040000
+#define B2R2_VMX1_YUV_TO_BLT_YUV888 0x00000100
+#define B2R2_VMX2_YUV_TO_BLT_YUV888 0x20000000
+#define B2R2_VMX3_YUV_TO_BLT_YUV888 0x00000000
+
+/* VMX register values for BLT_YUV888 to YUV conversion */
+#define B2R2_VMX0_BLT_YUV888_TO_YUV 0x00000100
+#define B2R2_VMX1_BLT_YUV888_TO_YUV 0x20000000
+#define B2R2_VMX2_BLT_YUV888_TO_YUV 0x00040000
+#define B2R2_VMX3_BLT_YUV888_TO_YUV 0x00000000
+
+/* VMX register values for YVU to BLT_YUV888 conversion */
+#define B2R2_VMX0_YVU_TO_BLT_YUV888 0x00040000
+#define B2R2_VMX1_YVU_TO_BLT_YUV888 0x20000000
+#define B2R2_VMX2_YVU_TO_BLT_YUV888 0x00000100
+#define B2R2_VMX3_YVU_TO_BLT_YUV888 0x00000000
+
+/* VMX register values for BLT_YUV888 to YVU conversion */
+#define B2R2_VMX0_BLT_YUV888_TO_YVU 0x00040000
+#define B2R2_VMX1_BLT_YUV888_TO_YVU 0x20000000
+#define B2R2_VMX2_BLT_YUV888_TO_YVU 0x00000100
+#define B2R2_VMX3_BLT_YUV888_TO_YVU 0x00000000
+
 #endif /* B2R2_HW_H__ */
diff --git a/drivers/video/b2r2/b2r2_input_validation.c b/drivers/video/b2r2/b2r2_input_validation.c
index 2604d8f77e0..0de714f1161 100644
--- a/drivers/video/b2r2/b2r2_input_validation.c
+++ b/drivers/video/b2r2/b2r2_input_validation.c
@@ -70,6 +70,8 @@ static bool is_valid_format(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
 	case B2R2_BLT_FMT_YUV444_PACKED_PLANAR:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		return true;
 
 	default:
@@ -105,9 +107,11 @@ static bool is_valid_pitch_for_fmt(u32 pitch, s32 width, enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		if (!b2r2_is_aligned(pitch, 4))
 			return false;
 
@@ -127,6 +131,7 @@ static bool is_aligned_width_for_fmt(s32 width, enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 		if (!b2r2_is_aligned(width, 4))
 			return false;
 
diff --git a/drivers/video/b2r2/b2r2_node_split.c b/drivers/video/b2r2/b2r2_node_split.c
index 90674e0e540..5aa62be8b4b 100644
--- a/drivers/video/b2r2/b2r2_node_split.c
+++ b/drivers/video/b2r2/b2r2_node_split.c
@@ -67,6 +67,33 @@ static const u32 vmx_blt_yuv888_to_rgb[] = {
 	B2R2_VMX3_BLT_YUV888_TO_RGB_601_VIDEO,
 };
 
+static const u32 vmx_yuv_to_blt_yuv888[] = {
+	B2R2_VMX0_YUV_TO_BLT_YUV888,
+	B2R2_VMX1_YUV_TO_BLT_YUV888,
+	B2R2_VMX2_YUV_TO_BLT_YUV888,
+	B2R2_VMX3_YUV_TO_BLT_YUV888,
+};
+
+static const u32 vmx_blt_yuv888_to_yuv[] = {
+	B2R2_VMX0_BLT_YUV888_TO_YUV,
+	B2R2_VMX1_BLT_YUV888_TO_YUV,
+	B2R2_VMX2_BLT_YUV888_TO_YUV,
+	B2R2_VMX3_BLT_YUV888_TO_YUV,
+};
+
+static const u32 vmx_yvu_to_blt_yuv888[] = {
+	B2R2_VMX0_YVU_TO_BLT_YUV888,
+	B2R2_VMX1_YVU_TO_BLT_YUV888,
+	B2R2_VMX2_YVU_TO_BLT_YUV888,
+	B2R2_VMX3_YVU_TO_BLT_YUV888,
+};
+
+static const u32 vmx_blt_yuv888_to_yvu[] = {
+	B2R2_VMX0_BLT_YUV888_TO_YVU,
+	B2R2_VMX1_BLT_YUV888_TO_YVU,
+	B2R2_VMX2_BLT_YUV888_TO_YVU,
+	B2R2_VMX3_BLT_YUV888_TO_YVU,
+};
 static const u32 vmx_yvu_to_rgb[] = {
 	B2R2_VMX0_YVU_TO_RGB_601_VIDEO,
 	B2R2_VMX1_YVU_TO_RGB_601_VIDEO,
@@ -270,6 +297,8 @@ int b2r2_node_split_analyze(const struct b2r2_blt_request *req,
 	switch (req->user_req.src_img.fmt) {
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		if (!is_rgb_fmt(req->user_req.dst_img.fmt)) {
 			ret = -ENOSYS;
 			goto unsupported;
@@ -283,6 +312,8 @@ int b2r2_node_split_analyze(const struct b2r2_blt_request *req,
 	switch (req->user_req.dst_img.fmt) {
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		if (!is_rgb_fmt(req->user_req.src_img.fmt)) {
 			ret = -ENOSYS;
 			goto unsupported;
@@ -692,7 +723,14 @@ static int analyze_fmt_conv(struct b2r2_node_split_buf *src,
 		if (is_yvu_fmt(dst->fmt))
 			*vmx = &vmx_rgb_to_yvu[0];
 		else if (dst->fmt == B2R2_BLT_FMT_24_BIT_YUV888 ||
-				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888)
+				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888 ||
+				dst->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+				dst->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+			/*
+			 * (A)YUV/VUY(A) formats differ only in component
+			 * order. This is handled by the endianness bit
+			 * in B2R2_STY/TTY registers when src/target are set.
+			 */
 			*vmx = &vmx_rgb_to_blt_yuv888[0];
 		else if (is_yuv_fmt(dst->fmt))
 			*vmx = &vmx_rgb_to_yuv[0];
@@ -704,29 +742,48 @@ static int analyze_fmt_conv(struct b2r2_node_split_buf *src,
 		else if (is_bgr_fmt(dst->fmt))
 			*vmx = &vmx_yvu_to_bgr[0];
 		else if (dst->fmt == B2R2_BLT_FMT_24_BIT_YUV888 ||
-				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888)
-			BUG_ON(1);
+				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888 ||
+				dst->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+				dst->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+			*vmx = &vmx_yvu_to_blt_yuv888[0];
 		else if (is_yuv_fmt(dst->fmt) &&
 				!is_yvu_fmt(dst->fmt))
 			*vmx = &vmx_yvu_to_yuv[0];
 	} else if (src->fmt == B2R2_BLT_FMT_24_BIT_YUV888 ||
-			src->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888) {
-		if (is_rgb_fmt(dst->fmt))
+			src->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888 ||
+			src->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+			src->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888) {
+		/*
+		 * (A)YUV/VUY(A) formats differ only in component
+		 * order. This is handled by the endianness bit
+		 * in B2R2_STY/TTY registers when src/target are set.
+		 */
+		if (is_rgb_fmt(dst->fmt)) {
 			*vmx = &vmx_blt_yuv888_to_rgb[0];
-		else
-			/*
-			 * not supported, b2r2_node_split_analyze()
-			 * should have returned ENOSYS.
-			 */
-			BUG_ON(1);
+		} else if (is_yvu_fmt(dst->fmt)) {
+			*vmx = &vmx_blt_yuv888_to_yvu[0];
+		} else if (is_yuv_fmt(dst->fmt)) {
+			switch (dst->fmt) {
+			case B2R2_BLT_FMT_24_BIT_YUV888:
+			case B2R2_BLT_FMT_32_BIT_AYUV8888:
+			case B2R2_BLT_FMT_24_BIT_VUY888:
+			case B2R2_BLT_FMT_32_BIT_VUYA8888: /* do nothing */
+				break;
+			default:
+				*vmx = &vmx_blt_yuv888_to_yuv[0];
+				break;
+			}
+		}
 	} else if (is_yuv_fmt(src->fmt)) {
 		if (is_rgb_fmt(dst->fmt))
 			*vmx = &vmx_yuv_to_rgb[0];
 		else if (is_bgr_fmt(dst->fmt))
 			*vmx = &vmx_yuv_to_bgr[0];
 		else if (dst->fmt == B2R2_BLT_FMT_24_BIT_YUV888 ||
-				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888)
-			BUG_ON(1);
+				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888 ||
+				dst->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+				dst->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+			*vmx = &vmx_yuv_to_blt_yuv888[0];
 		else if (is_yvu_fmt(dst->fmt))
 			*vmx = &vmx_yvu_to_yuv[0];
 	} else if (is_bgr_fmt(src->fmt)) {
@@ -735,7 +792,9 @@ static int analyze_fmt_conv(struct b2r2_node_split_buf *src,
 		else if (is_yvu_fmt(dst->fmt))
 			*vmx = &vmx_bgr_to_yvu[0];
 		else if (dst->fmt == B2R2_BLT_FMT_24_BIT_YUV888 ||
-				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888)
+				dst->fmt == B2R2_BLT_FMT_32_BIT_AYUV8888 ||
+				dst->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+				dst->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
 			BUG_ON(1);
 		else if (is_yuv_fmt(dst->fmt))
 			*vmx = &vmx_bgr_to_yuv[0];
@@ -778,7 +837,8 @@ static int analyze_color_fill(struct b2r2_node_split_job *this,
 	if ((!this->blend) && ((this->flags & B2R2_BLT_FLAG_SOURCE_FILL_RAW) ||
 			(this->dst.fmt == B2R2_BLT_FMT_32_BIT_ARGB8888) ||
 			(this->dst.fmt == B2R2_BLT_FMT_32_BIT_ABGR8888) ||
-			(this->dst.fmt == B2R2_BLT_FMT_32_BIT_AYUV8888))) {
+			(this->dst.fmt == B2R2_BLT_FMT_32_BIT_AYUV8888) ||
+			(this->dst.fmt == B2R2_BLT_FMT_32_BIT_VUYA8888))) {
 		this->type = B2R2_DIRECT_FILL;
 
 		/* The color format will be the same as the dst fmt */
@@ -2707,6 +2767,7 @@ static enum b2r2_ty get_alpha_range(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_8_BIT_A8:
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 		return B2R2_TY_ALPHA_RANGE_255; /* 0 - 255 */
@@ -2725,6 +2786,8 @@ static u8 get_alpha(enum b2r2_blt_fmt fmt, u32 pixel)
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
 		return (pixel >> 24) & 0xff;
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
+		return pixel & 0xff;
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 		return (pixel & 0xfff) >> 16;
 	case B2R2_BLT_FMT_16_BIT_ARGB4444:
@@ -2754,6 +2817,10 @@ static u32 set_alpha(enum b2r2_blt_fmt fmt, u8 alpha, u32 color)
 		color &= 0x00ffffff;
 		alpha_mask = alpha << 24;
 		break;
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
+		color &= 0xffffff00;
+		alpha_mask = alpha;
+		break;
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 		color &= 0x00ffff;
 		alpha_mask = alpha << 16;
@@ -2793,6 +2860,7 @@ static bool fmt_has_alpha(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_1_BIT_A1:
 	case B2R2_BLT_FMT_8_BIT_A8:
 		return true;
@@ -2837,6 +2905,8 @@ static bool is_yuv_fmt(enum b2r2_blt_fmt fmt)
 	switch (fmt) {
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_Y_CB_Y_CR:
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
 	case B2R2_BLT_FMT_YUV420_PACKED_PLANAR:
@@ -2942,6 +3012,8 @@ static bool is_yuv444_fmt(enum b2r2_blt_fmt fmt)
 	switch (fmt) {
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_YUV444_PACKED_PLANAR:
 		return true;
 	default:
@@ -2988,11 +3060,13 @@ static int fmt_byte_pitch(enum b2r2_blt_fmt fmt, u32 width)
 	case B2R2_BLT_FMT_24_BIT_RGB888:   /* Fall through */
 	case B2R2_BLT_FMT_24_BIT_ARGB8565: /* Fall through */
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 		return width * 3;
 
 	case B2R2_BLT_FMT_32_BIT_ARGB8888: /* Fall through */
 	case B2R2_BLT_FMT_32_BIT_ABGR8888: /* Fall through */
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		return width << 2;
 
 	default:
@@ -3026,11 +3100,13 @@ static enum b2r2_native_fmt to_native_fmt(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 		return B2R2_NATIVE_RGB888;
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888: /* Not actually supported by HW */
 		return B2R2_NATIVE_YCBCR888;
 	case B2R2_BLT_FMT_32_BIT_ABGR8888: /* Not actually supported by HW */
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 		return B2R2_NATIVE_ARGB8888;
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888: /* Not actually supported by HW */
 		return B2R2_NATIVE_AYCBCR8888;
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
 		return B2R2_NATIVE_YCBCR422R;
@@ -3125,6 +3201,8 @@ static enum b2r2_fmt_type get_fmt_type(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_1_BIT_A1:
 	case B2R2_BLT_FMT_8_BIT_A8:
 		return B2R2_FMT_TYPE_RASTER;
@@ -3256,6 +3334,11 @@ static void set_target(struct b2r2_node *node, u32 addr,
 	node->node.GROUP1.B2R2_TTY = buf->pitch | to_native_fmt(buf->fmt) |
 			buf->alpha_range | buf->chroma_selection | buf->hso |
 			buf->vso | buf->dither | buf->plane_selection;
+
+	if (buf->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+			buf->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+		node->node.GROUP1.B2R2_TTY |= B2R2_TY_ENDIAN_BIG_NOT_LITTLE;
+
 	node->node.GROUP1.B2R2_TSZ =
 			((buf->win.width & 0xfff) << B2R2_SZ_WIDTH_SHIFT) |
 			((buf->win.height & 0xfff) << B2R2_SZ_HEIGHT_SHIFT);
@@ -3306,6 +3389,11 @@ static void set_src(struct b2r2_src_config *src, u32 addr,
 	src->B2R2_SBA = addr;
 	src->B2R2_STY = buf->pitch | to_native_fmt(buf->fmt) |
 			buf->alpha_range | buf->hso | buf->vso;
+
+	if (buf->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+			buf->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+		src->B2R2_STY |= B2R2_TY_ENDIAN_BIG_NOT_LITTLE;
+
 	src->B2R2_SSZ = ((buf->win.width & 0xfff) << B2R2_SZ_WIDTH_SHIFT) |
 			((buf->win.height & 0xfff) << B2R2_SZ_HEIGHT_SHIFT);
 	src->B2R2_SXY = ((buf->win.x & 0xffff) << B2R2_XY_X_SHIFT) |
@@ -3330,6 +3418,11 @@ static void set_src_1(struct b2r2_node *node, u32 addr,
 	node->node.GROUP3.B2R2_SBA = addr;
 	node->node.GROUP3.B2R2_STY = buf->pitch | to_native_fmt(buf->fmt) |
 			buf->alpha_range | buf->hso | buf->vso;
+
+	if (buf->fmt == B2R2_BLT_FMT_24_BIT_VUY888 ||
+			buf->fmt == B2R2_BLT_FMT_32_BIT_VUYA8888)
+		node->node.GROUP3.B2R2_STY |= B2R2_TY_ENDIAN_BIG_NOT_LITTLE;
+
 	node->node.GROUP3.B2R2_SXY =
 			((buf->win.x & 0xffff) << B2R2_XY_X_SHIFT) |
 			((buf->win.y & 0xffff) << B2R2_XY_Y_SHIFT);
diff --git a/drivers/video/b2r2/b2r2_utils.c b/drivers/video/b2r2/b2r2_utils.c
index 44ea06e51fd..ad0e7a42fa1 100644
--- a/drivers/video/b2r2/b2r2_utils.c
+++ b/drivers/video/b2r2/b2r2_utils.c
@@ -101,12 +101,14 @@ int b2r2_get_fmt_bpp(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 	case B2R2_BLT_FMT_YUV444_PACKED_PLANAR:
 		return 24;
 
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		return 32;
 
 	default:
@@ -134,6 +136,8 @@ int b2r2_get_fmt_y_bpp(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_YUV444_PACKED_PLANAR:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		return 8;
 
 	default:
@@ -155,9 +159,11 @@ bool b2r2_is_single_plane_fmt(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_Y_CB_Y_CR:
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
 		return true;
@@ -178,9 +184,11 @@ bool b2r2_is_independent_pixel_fmt(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_24_BIT_RGB888:
 	case B2R2_BLT_FMT_24_BIT_ARGB8565:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 	case B2R2_BLT_FMT_32_BIT_ARGB8888:
 	case B2R2_BLT_FMT_32_BIT_ABGR8888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 	case B2R2_BLT_FMT_YUV444_PACKED_PLANAR:
 		return true;
 
@@ -196,6 +204,8 @@ bool b2r2_is_ycbcri_fmt(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_CB_Y_CR_Y:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
 		return true;
 
 	default:
@@ -270,6 +280,8 @@ bool b2r2_is_ycbcr444_fmt(enum b2r2_blt_fmt fmt)
 	case B2R2_BLT_FMT_YUV444_PACKED_PLANAR:
 	case B2R2_BLT_FMT_32_BIT_AYUV8888:
 	case B2R2_BLT_FMT_24_BIT_YUV888:
+	case B2R2_BLT_FMT_32_BIT_VUYA8888:
+	case B2R2_BLT_FMT_24_BIT_VUY888:
 		return true;
 
 	default:
diff --git a/include/video/b2r2_blt.h b/include/video/b2r2_blt.h
index 1cf5f661598..a992155f0da 100644
--- a/include/video/b2r2_blt.h
+++ b/include/video/b2r2_blt.h
@@ -127,6 +127,10 @@ struct b2r2_blt_rect {
  * The buffer shall contain a plane of Y, V, and U data in this order.
  * (Same as B2R2_BLT_FMT_YUV422_PACKED_PLANAR except that chroma
  * order is swapped.)
+ * @B2R2_BLT_FMT_24_BIT_VUY888: 24 bits per pixel VUY format with colors
+ * stored as V 23:16, U 15:8, and Y 7:0.
+ * @B2R2_BLT_FMT_32_BIT_VUYA8888: 32 bits per pixel VUYA format with colors
+ * stored as V 31:24, U 23:16, Y 15:8, and Alpha 7:0.
  */
 enum b2r2_blt_fmt {
     B2R2_BLT_FMT_UNUSED = 0,
@@ -155,6 +159,8 @@ enum b2r2_blt_fmt {
     B2R2_BLT_FMT_YVU422_PACKED_SEMI_PLANAR = 0x7F00000A,
     B2R2_BLT_FMT_YVU420_PACKED_PLANAR = 0x7F00000B,
     B2R2_BLT_FMT_YVU422_PACKED_PLANAR = 0x7F00000C,
+    B2R2_BLT_FMT_24_BIT_VUY888 = 0x7F00000D,
+    B2R2_BLT_FMT_32_BIT_VUYA8888 = 0x7F00000E,
 };
 
 /**