[PATCH v2 00/13] media: rockchip: rga: add support for multi-planar formats

[PATCH v2 00/13] media: rockchip: rga: add support for multi-planar formats

[Michael Tretter]

This is v2 of the series that adds support for the V4L2 multi-planar API
to the Rockchip RGA driver. Once the RGA driver supports the
multi-planar API, it is easier to share buffers with other V4L2 drivers
that also support the multi-planar API and may not expose planar formats
with contiguous planes.

v2 fixes the smatch warnings and compile errors of v1. Furthermore, now
the DMA mask is set to 32 bits for coherent, too, and the gfp_flags are
configured to ensure that buffers are allocated in the lower 4GB area.

With non-contiguous planes, the U and V planes may not start at the same
offset as with the continuous planes. Therefore, the RGA driver cannot
rely on its calculation of the plane offsets based on the format and
frame size anymore, but must remember the offsets when it created the
mapping. Therefore, I also reworked how the DMA mapping is handled.

As a bonus, the RGA driver should now work correctly on devices with
more than 4 GB of memory. Video buffers should now be allocated within
the 4 GB boundary and an import of buffers that have higher addresses
into the driver should fail.

Patch 1 fixes the swizzling of RGA formats. While testing all formats of
the driver on rk3568, I discovered that the color channels of the RGB
formats are wrong when converting to NV12. I didn't test this on other
SoCs with an RGA and I am not sure, if they behave differently regarding
the color channels. Please report, if this breaks the color conversion
on other SoCs, and I will make this SoC-specific.

Patches 2 to 6 are the rework the DMA descriptor handling for the RGA
MMU. The patches clean up, how the driver uses the DMA API, and make the
creation of the descriptor list more explicit. Furthermore, the driver
is changed to keep the mapping per video buffer instead of using a
single mapping that is updated with every buffer.

Patches 7 to 11 prepare the driver for the multi-planar API including a
cleanup of the format handling in the buffer, and finally switch to the
multi-planar API.

Patch 12 updates the code that creates the DMA-descriptor mapping to
correctly handle buffers with multiple planes. The driver has to iterate
all planes and make them the continuous for the RGA.

Patch 13 enables the NV12M format, which is the multi-planar variant of
the NV12 format.

Michael

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2:
- Fix smatch warnings
- Fix cast to dst_mmu_pages/src_mmu_pages to fix compile error in Patch 2
- Remove check for upper_32_bits when filling the DMA descriptor table
- Remove useless dma_sync_single_for_device()
- Set DMA mask for DMA coherent
- Set gfp_flags to __GFP_DMA32
- Link to v1: https://lore.kernel.org/r/20230914-rockchip-rga-multiplanar-v1-0-abfd77260ae3@pengutronix.de

---
Michael Tretter (13):
      media: rockchip: rga: fix swizzling for RGB formats
      media: rockchip: rga: extract helper to fill descriptors
      media: rockchip: rga: allocate DMA descriptors per buffer
      media: rockchip: rga: split src and dst buffer setup
      media: rockchip: rga: pre-calculate plane offsets
      media: rockchip: rga: set dma mask to 32 bits
      media: rockchip: rga: use clamp() to clamp size to limits
      media: rockchip: rga: use pixelformat to find format
      media: rockchip: rga: add local variable for pix_format
      media: rockchip: rga: use macros for testing buffer type
      media: rockchip: rga: switch to multi-planar API
      media: rockchip: rga: rework buffer handling for multi-planar formats
      media: rockchip: rga: add NV12M support

 drivers/media/platform/rockchip/rga/rga-buf.c | 162 ++++++++++++++++------
 drivers/media/platform/rockchip/rga/rga-hw.c  | 146 ++++++++++++--------
 drivers/media/platform/rockchip/rga/rga.c     | 189 ++++++++++++--------------
 drivers/media/platform/rockchip/rga/rga.h     |  35 ++++-
 4 files changed, 328 insertions(+), 204 deletions(-)
---
base-commit: 2c1bae27df787c9535e48cc27bbd11c3c3e0a235
change-id: 20230914-rockchip-rga-multiplanar-0d7c79b1ba93

Best regards,
-- 
Michael Tretter <m.tretter@pengutronix.de>


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 01/13] media: rockchip: rga: fix swizzling for RGB formats

[Michael Tretter]

When using 32 bit RGB formats, the RGA on the rk3568 produces wrong
colors as the wrong color channels are read or written.  The reason is
that the format description for the channel swizzeling is wrong and the
wrong bits are configured. For example, when converting ARGB32 to NV12,
the alpha channel is used as blue channel.. This doesn't happen if the
color format is the same on both sides.

Fix the color_swap settings of the formats to correctly handle 32 bit
RGB formats.

For RGA_COLOR_FMT_XBGR8888, the RGA_COLOR_ALPHA_SWAP bit doesn't have an
effect. Thus, it isn't possible to handle the V4L2_PIX_FMT_XRGB32. Thus,
it is removed from the list of supported formats.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 15 +++------------
 1 file changed, 3 insertions(+), 12 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index f1c532a5802a..25f5b5eebf13 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -184,25 +184,16 @@ static int rga_setup_ctrls(struct rga_ctx *ctx)
 static struct rga_fmt formats[] = {
 	{
 		.fourcc = V4L2_PIX_FMT_ARGB32,
-		.color_swap = RGA_COLOR_RB_SWAP,
+		.color_swap = RGA_COLOR_ALPHA_SWAP,
 		.hw_format = RGA_COLOR_FMT_ABGR8888,
 		.depth = 32,
 		.uv_factor = 1,
 		.y_div = 1,
 		.x_div = 1,
 	},
-	{
-		.fourcc = V4L2_PIX_FMT_XRGB32,
-		.color_swap = RGA_COLOR_RB_SWAP,
-		.hw_format = RGA_COLOR_FMT_XBGR8888,
-		.depth = 32,
-		.uv_factor = 1,
-		.y_div = 1,
-		.x_div = 1,
-	},
 	{
 		.fourcc = V4L2_PIX_FMT_ABGR32,
-		.color_swap = RGA_COLOR_ALPHA_SWAP,
+		.color_swap = RGA_COLOR_RB_SWAP,
 		.hw_format = RGA_COLOR_FMT_ABGR8888,
 		.depth = 32,
 		.uv_factor = 1,
@@ -211,7 +202,7 @@ static struct rga_fmt formats[] = {
 	},
 	{
 		.fourcc = V4L2_PIX_FMT_XBGR32,
-		.color_swap = RGA_COLOR_ALPHA_SWAP,
+		.color_swap = RGA_COLOR_RB_SWAP,
 		.hw_format = RGA_COLOR_FMT_XBGR8888,
 		.depth = 32,
 		.uv_factor = 1,

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 01/13] media: rockchip: rga: fix swizzling for RGB formats

[Deborah Brouwer]

On Fri, Oct 13, 2023 at 01:00:22PM +0200, Michael Tretter wrote:

Hi Michael,

> When using 32 bit RGB formats, the RGA on the rk3568 produces wrong
> colors as the wrong color channels are read or written.  The reason is
> that the format description for the channel swizzeling is wrong and the
> wrong bits are configured. For example, when converting ARGB32 to NV12,
> the alpha channel is used as blue channel.. This doesn't happen if the
> color format is the same on both sides.
> 
> Fix the color_swap settings of the formats to correctly handle 32 bit
> RGB formats.

You mentioned in your cover letter that you only tested this on rk3568
so I tested it on an rk3399 as well.

I used gstreamer to convert each of the formats that you changed
(V4L2_PIX_FMT_ARGB32, V4L2_PIX_FMT_ABGR32, V4L2_PIX_FMT_XBGR32) into
each of:
BA24, AR24, XR24, RGB3, BGRI 3, NV21, YU12, 422P, YV12, and NM12.

All the colour conversions are working and much better than before which
I could see was definitely broken.

Tested-by: Deborah Brouwer <deborah.brouwer@collabora.com>

> 
> For RGA_COLOR_FMT_XBGR8888, the RGA_COLOR_ALPHA_SWAP bit doesn't have an
> effect. Thus, it isn't possible to handle the V4L2_PIX_FMT_XRGB32. Thus,
> it is removed from the list of supported formats.
> 
> Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
> ---
> Changes in v2: None
> ---
>  drivers/media/platform/rockchip/rga/rga.c | 15 +++------------
>  1 file changed, 3 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
> index f1c532a5802a..25f5b5eebf13 100644
> --- a/drivers/media/platform/rockchip/rga/rga.c
> +++ b/drivers/media/platform/rockchip/rga/rga.c
> @@ -184,25 +184,16 @@ static int rga_setup_ctrls(struct rga_ctx *ctx)
>  static struct rga_fmt formats[] = {
>  	{
>  		.fourcc = V4L2_PIX_FMT_ARGB32,
> -		.color_swap = RGA_COLOR_RB_SWAP,
> +		.color_swap = RGA_COLOR_ALPHA_SWAP,
>  		.hw_format = RGA_COLOR_FMT_ABGR8888,
>  		.depth = 32,
>  		.uv_factor = 1,
>  		.y_div = 1,
>  		.x_div = 1,
>  	},
> -	{
> -		.fourcc = V4L2_PIX_FMT_XRGB32,
> -		.color_swap = RGA_COLOR_RB_SWAP,
> -		.hw_format = RGA_COLOR_FMT_XBGR8888,
> -		.depth = 32,
> -		.uv_factor = 1,
> -		.y_div = 1,
> -		.x_div = 1,
> -	},
>  	{
>  		.fourcc = V4L2_PIX_FMT_ABGR32,
> -		.color_swap = RGA_COLOR_ALPHA_SWAP,
> +		.color_swap = RGA_COLOR_RB_SWAP,
>  		.hw_format = RGA_COLOR_FMT_ABGR8888,
>  		.depth = 32,
>  		.uv_factor = 1,
> @@ -211,7 +202,7 @@ static struct rga_fmt formats[] = {
>  	},
>  	{
>  		.fourcc = V4L2_PIX_FMT_XBGR32,
> -		.color_swap = RGA_COLOR_ALPHA_SWAP,
> +		.color_swap = RGA_COLOR_RB_SWAP,
>  		.hw_format = RGA_COLOR_FMT_XBGR8888,
>  		.depth = 32,
>  		.uv_factor = 1,
> 
> -- 
> 2.39.2
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 02/13] media: rockchip: rga: extract helper to fill descriptors

[Michael Tretter]

The IOMMU of the RGA is programmed with a list of DMA descriptors that
contain an 32 bit address per 4k page in the video buffers. The address
in the descriptor points to the start address of the page.

Introduce 'struct rga_dma_desc' to make the handling of the DMA
descriptors explicit instead of hiding them behind standard types.

While at it, use provided helpers for iterating the sg_table instead of
manually calculating the DMA addresses.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2:
- Remove check for upper_32_bits()
- Fix cast to dst_mmu_pages/src_mmu_pages
---
 drivers/media/platform/rockchip/rga/rga-buf.c | 43 ++++++++++++++-------------
 drivers/media/platform/rockchip/rga/rga.c     |  4 +--
 drivers/media/platform/rockchip/rga/rga.h     |  8 +++--
 3 files changed, 30 insertions(+), 25 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga-buf.c b/drivers/media/platform/rockchip/rga/rga-buf.c
index 81508ed5abf3..fc3f38f6b5e1 100644
--- a/drivers/media/platform/rockchip/rga/rga-buf.c
+++ b/drivers/media/platform/rockchip/rga/rga-buf.c
@@ -5,6 +5,7 @@
  */
 
 #include <linux/pm_runtime.h>
+#include <linux/scatterlist.h>
 
 #include <media/v4l2-device.h>
 #include <media/v4l2-ioctl.h>
@@ -15,6 +16,23 @@
 #include "rga-hw.h"
 #include "rga.h"
 
+static size_t fill_descriptors(struct rga_dma_desc *desc, struct sg_table *sgt)
+{
+	struct sg_dma_page_iter iter;
+	struct rga_dma_desc *tmp = desc;
+	size_t n_desc = 0;
+	dma_addr_t addr;
+
+	for_each_sgtable_dma_page(sgt, &iter, 0) {
+		addr = sg_page_iter_dma_address(&iter);
+		tmp->addr = lower_32_bits(addr);
+		tmp++;
+		n_desc++;
+	}
+
+	return n_desc;
+}
+
 static int
 rga_queue_setup(struct vb2_queue *vq,
 		unsigned int *nbuffers, unsigned int *nplanes,
@@ -114,11 +132,8 @@ void rga_buf_map(struct vb2_buffer *vb)
 {
 	struct rga_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
 	struct rockchip_rga *rga = ctx->rga;
-	struct sg_table *sgt;
-	struct scatterlist *sgl;
-	unsigned int *pages;
-	unsigned int address, len, i, p;
-	unsigned int mapped_size = 0;
+	struct rga_dma_desc *pages;
+	size_t n_desc = 0;
 
 	if (vb->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
 		pages = rga->src_mmu_pages;
@@ -126,23 +141,9 @@ void rga_buf_map(struct vb2_buffer *vb)
 		pages = rga->dst_mmu_pages;
 
 	/* Create local MMU table for RGA */
-	sgt = vb2_plane_cookie(vb, 0);
-
-	for_each_sg(sgt->sgl, sgl, sgt->nents, i) {
-		len = sg_dma_len(sgl) >> PAGE_SHIFT;
-		address = sg_phys(sgl);
-
-		for (p = 0; p < len; p++) {
-			dma_addr_t phys = address +
-					  ((dma_addr_t)p << PAGE_SHIFT);
-
-			pages[mapped_size + p] = phys;
-		}
-
-		mapped_size += len;
-	}
+	n_desc = fill_descriptors(pages, vb2_dma_sg_plane_desc(vb, 0));
 
 	/* sync local MMU table for RGA */
 	dma_sync_single_for_device(rga->dev, virt_to_phys(pages),
-				   8 * PAGE_SIZE, DMA_BIDIRECTIONAL);
+				   n_desc * sizeof(*pages), DMA_BIDIRECTIONAL);
 }
diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index 25f5b5eebf13..36d7f255af90 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -873,13 +873,13 @@ static int rga_probe(struct platform_device *pdev)
 	}
 
 	rga->src_mmu_pages =
-		(unsigned int *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 3);
+		(struct rga_dma_desc *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 3);
 	if (!rga->src_mmu_pages) {
 		ret = -ENOMEM;
 		goto free_dma;
 	}
 	rga->dst_mmu_pages =
-		(unsigned int *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 3);
+		(struct rga_dma_desc *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 3);
 	if (!rga->dst_mmu_pages) {
 		ret = -ENOMEM;
 		goto free_src_pages;
diff --git a/drivers/media/platform/rockchip/rga/rga.h b/drivers/media/platform/rockchip/rga/rga.h
index 5fa9d2f366dc..22f7da28ac51 100644
--- a/drivers/media/platform/rockchip/rga/rga.h
+++ b/drivers/media/platform/rockchip/rga/rga.h
@@ -40,6 +40,10 @@ struct rga_frame {
 	u32 size;
 };
 
+struct rga_dma_desc {
+	u32 addr;
+};
+
 struct rockchip_rga_version {
 	u32 major;
 	u32 minor;
@@ -81,8 +85,8 @@ struct rockchip_rga {
 	struct rga_ctx *curr;
 	dma_addr_t cmdbuf_phy;
 	void *cmdbuf_virt;
-	unsigned int *src_mmu_pages;
-	unsigned int *dst_mmu_pages;
+	struct rga_dma_desc *src_mmu_pages;
+	struct rga_dma_desc *dst_mmu_pages;
 };
 
 struct rga_frame *rga_get_frame(struct rga_ctx *ctx, enum v4l2_buf_type type);

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 03/13] media: rockchip: rga: allocate DMA descriptors per buffer

[Michael Tretter]

The RGA driver allocates two buffers for the DMA descriptors of the
input and output buffers. Whenever a new job is processed, the
descriptor list is updated for the current buffers.

By updating the descriptor list during buf_prepare, it is possible to
correctly fail DMABUF imports if the buffers that shall be imported are
not within the 32 bit address range that can be addressed by the RGA.

Managing the DMA descriptor list with the buffer also makes it easier to
track the buffer mapping and the plane offsets into this mapping.

The cost is that the driver now requires DMA coherent memory per buffer
for the descriptor list. However, the size scales with the size of the
video buffers and is not allocated if the RGA is not used.

While at it, use dma_alloc_coherent to allocate the descriptors and get
rid of the virt_to_phys calls to get the physical addresses.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2:
- Drop useless dma_sync_single_for_device()
---
 drivers/media/platform/rockchip/rga/rga-buf.c | 62 +++++++++++++++++----------
 drivers/media/platform/rockchip/rga/rga-hw.c  | 26 +++++------
 drivers/media/platform/rockchip/rga/rga.c     | 31 ++------------
 drivers/media/platform/rockchip/rga/rga.h     | 21 +++++++--
 4 files changed, 74 insertions(+), 66 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga-buf.c b/drivers/media/platform/rockchip/rga/rga-buf.c
index fc3f38f6b5e1..ff36efade060 100644
--- a/drivers/media/platform/rockchip/rga/rga-buf.c
+++ b/drivers/media/platform/rockchip/rga/rga-buf.c
@@ -53,8 +53,31 @@ rga_queue_setup(struct vb2_queue *vq,
 	return 0;
 }
 
+static int rga_buf_init(struct vb2_buffer *vb)
+{
+	struct vb2_v4l2_buffer *vbuf = to_vb2_v4l2_buffer(vb);
+	struct rga_vb_buffer *rbuf = vb_to_rga(vbuf);
+	struct rga_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
+	struct rockchip_rga *rga = ctx->rga;
+	struct rga_frame *f = rga_get_frame(ctx, vb->vb2_queue->type);
+	size_t n_desc = 0;
+
+	n_desc = DIV_ROUND_UP(f->size, PAGE_SIZE);
+
+	rbuf->n_desc = n_desc;
+	rbuf->dma_desc = dma_alloc_coherent(rga->dev,
+					    rbuf->n_desc * sizeof(*rbuf->dma_desc),
+					    &rbuf->dma_desc_pa, GFP_KERNEL);
+	if (!rbuf->dma_desc)
+		return -ENOMEM;
+
+	return 0;
+}
+
 static int rga_buf_prepare(struct vb2_buffer *vb)
 {
+	struct vb2_v4l2_buffer *vbuf = to_vb2_v4l2_buffer(vb);
+	struct rga_vb_buffer *rbuf = vb_to_rga(vbuf);
 	struct rga_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
 	struct rga_frame *f = rga_get_frame(ctx, vb->vb2_queue->type);
 
@@ -63,6 +86,9 @@ static int rga_buf_prepare(struct vb2_buffer *vb)
 
 	vb2_set_plane_payload(vb, 0, f->size);
 
+	/* Create local MMU table for RGA */
+	fill_descriptors(rbuf->dma_desc, vb2_dma_sg_plane_desc(vb, 0));
+
 	return 0;
 }
 
@@ -74,6 +100,17 @@ static void rga_buf_queue(struct vb2_buffer *vb)
 	v4l2_m2m_buf_queue(ctx->fh.m2m_ctx, vbuf);
 }
 
+static void rga_buf_cleanup(struct vb2_buffer *vb)
+{
+	struct vb2_v4l2_buffer *vbuf = to_vb2_v4l2_buffer(vb);
+	struct rga_vb_buffer *rbuf = vb_to_rga(vbuf);
+	struct rga_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
+	struct rockchip_rga *rga = ctx->rga;
+
+	dma_free_coherent(rga->dev, rbuf->n_desc * sizeof(*rbuf->dma_desc),
+			  rbuf->dma_desc, rbuf->dma_desc_pa);
+}
+
 static void rga_buf_return_buffers(struct vb2_queue *q,
 				   enum vb2_buffer_state state)
 {
@@ -117,33 +154,12 @@ static void rga_buf_stop_streaming(struct vb2_queue *q)
 
 const struct vb2_ops rga_qops = {
 	.queue_setup = rga_queue_setup,
+	.buf_init = rga_buf_init,
 	.buf_prepare = rga_buf_prepare,
 	.buf_queue = rga_buf_queue,
+	.buf_cleanup = rga_buf_cleanup,
 	.wait_prepare = vb2_ops_wait_prepare,
 	.wait_finish = vb2_ops_wait_finish,
 	.start_streaming = rga_buf_start_streaming,
 	.stop_streaming = rga_buf_stop_streaming,
 };
-
-/* RGA MMU is a 1-Level MMU, so it can't be used through the IOMMU API.
- * We use it more like a scatter-gather list.
- */
-void rga_buf_map(struct vb2_buffer *vb)
-{
-	struct rga_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
-	struct rockchip_rga *rga = ctx->rga;
-	struct rga_dma_desc *pages;
-	size_t n_desc = 0;
-
-	if (vb->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
-		pages = rga->src_mmu_pages;
-	else
-		pages = rga->dst_mmu_pages;
-
-	/* Create local MMU table for RGA */
-	n_desc = fill_descriptors(pages, vb2_dma_sg_plane_desc(vb, 0));
-
-	/* sync local MMU table for RGA */
-	dma_sync_single_for_device(rga->dev, virt_to_phys(pages),
-				   n_desc * sizeof(*pages), DMA_BIDIRECTIONAL);
-}
diff --git a/drivers/media/platform/rockchip/rga/rga-hw.c b/drivers/media/platform/rockchip/rga/rga-hw.c
index aaa96f256356..b391d97d4632 100644
--- a/drivers/media/platform/rockchip/rga/rga-hw.c
+++ b/drivers/media/platform/rockchip/rga/rga-hw.c
@@ -119,40 +119,40 @@ static struct rga_addr_offset *rga_lookup_draw_pos(struct
 	return NULL;
 }
 
-static void rga_cmd_set_src_addr(struct rga_ctx *ctx, void *mmu_pages)
+static void rga_cmd_set_src_addr(struct rga_ctx *ctx, dma_addr_t dma_addr)
 {
 	struct rockchip_rga *rga = ctx->rga;
 	u32 *dest = rga->cmdbuf_virt;
 	unsigned int reg;
 
 	reg = RGA_MMU_SRC_BASE - RGA_MODE_BASE_REG;
-	dest[reg >> 2] = virt_to_phys(mmu_pages) >> 4;
+	dest[reg >> 2] = dma_addr >> 4;
 
 	reg = RGA_MMU_CTRL1 - RGA_MODE_BASE_REG;
 	dest[reg >> 2] |= 0x7;
 }
 
-static void rga_cmd_set_src1_addr(struct rga_ctx *ctx, void *mmu_pages)
+static void rga_cmd_set_src1_addr(struct rga_ctx *ctx, dma_addr_t dma_addr)
 {
 	struct rockchip_rga *rga = ctx->rga;
 	u32 *dest = rga->cmdbuf_virt;
 	unsigned int reg;
 
 	reg = RGA_MMU_SRC1_BASE - RGA_MODE_BASE_REG;
-	dest[reg >> 2] = virt_to_phys(mmu_pages) >> 4;
+	dest[reg >> 2] = dma_addr >> 4;
 
 	reg = RGA_MMU_CTRL1 - RGA_MODE_BASE_REG;
 	dest[reg >> 2] |= 0x7 << 4;
 }
 
-static void rga_cmd_set_dst_addr(struct rga_ctx *ctx, void *mmu_pages)
+static void rga_cmd_set_dst_addr(struct rga_ctx *ctx, dma_addr_t dma_addr)
 {
 	struct rockchip_rga *rga = ctx->rga;
 	u32 *dest = rga->cmdbuf_virt;
 	unsigned int reg;
 
 	reg = RGA_MMU_DST_BASE - RGA_MODE_BASE_REG;
-	dest[reg >> 2] = virt_to_phys(mmu_pages) >> 4;
+	dest[reg >> 2] = dma_addr >> 4;
 
 	reg = RGA_MMU_CTRL1 - RGA_MODE_BASE_REG;
 	dest[reg >> 2] |= 0x7 << 8;
@@ -375,20 +375,21 @@ static void rga_cmd_set_mode(struct rga_ctx *ctx)
 	dest[(RGA_MODE_CTRL - RGA_MODE_BASE_REG) >> 2] = mode.val;
 }
 
-static void rga_cmd_set(struct rga_ctx *ctx)
+static void rga_cmd_set(struct rga_ctx *ctx,
+			struct rga_vb_buffer *src, struct rga_vb_buffer *dst)
 {
 	struct rockchip_rga *rga = ctx->rga;
 
 	memset(rga->cmdbuf_virt, 0, RGA_CMDBUF_SIZE * 4);
 
-	rga_cmd_set_src_addr(ctx, rga->src_mmu_pages);
+	rga_cmd_set_src_addr(ctx, src->dma_desc_pa);
 	/*
 	 * Due to hardware bug,
 	 * src1 mmu also should be configured when using alpha blending.
 	 */
-	rga_cmd_set_src1_addr(ctx, rga->dst_mmu_pages);
+	rga_cmd_set_src1_addr(ctx, dst->dma_desc_pa);
 
-	rga_cmd_set_dst_addr(ctx, rga->dst_mmu_pages);
+	rga_cmd_set_dst_addr(ctx, dst->dma_desc_pa);
 	rga_cmd_set_mode(ctx);
 
 	rga_cmd_set_trans_info(ctx);
@@ -400,11 +401,12 @@ static void rga_cmd_set(struct rga_ctx *ctx)
 		PAGE_SIZE, DMA_BIDIRECTIONAL);
 }
 
-void rga_hw_start(struct rockchip_rga *rga)
+void rga_hw_start(struct rockchip_rga *rga,
+		  struct rga_vb_buffer *src, struct rga_vb_buffer *dst)
 {
 	struct rga_ctx *ctx = rga->curr;
 
-	rga_cmd_set(ctx);
+	rga_cmd_set(ctx, src, dst);
 
 	rga_write(rga, RGA_SYS_CTRL, 0x00);
 
diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index 36d7f255af90..f18fccc7b204 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -45,10 +45,7 @@ static void device_run(void *prv)
 	src = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
 	dst = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
 
-	rga_buf_map(&src->vb2_buf);
-	rga_buf_map(&dst->vb2_buf);
-
-	rga_hw_start(rga);
+	rga_hw_start(rga, vb_to_rga(src), vb_to_rga(dst));
 
 	spin_unlock_irqrestore(&rga->ctrl_lock, flags);
 }
@@ -101,7 +98,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
 	src_vq->drv_priv = ctx;
 	src_vq->ops = &rga_qops;
 	src_vq->mem_ops = &vb2_dma_sg_memops;
-	src_vq->buf_struct_size = sizeof(struct v4l2_m2m_buffer);
+	src_vq->buf_struct_size = sizeof(struct rga_vb_buffer);
 	src_vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
 	src_vq->lock = &ctx->rga->mutex;
 	src_vq->dev = ctx->rga->v4l2_dev.dev;
@@ -115,7 +112,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
 	dst_vq->drv_priv = ctx;
 	dst_vq->ops = &rga_qops;
 	dst_vq->mem_ops = &vb2_dma_sg_memops;
-	dst_vq->buf_struct_size = sizeof(struct v4l2_m2m_buffer);
+	dst_vq->buf_struct_size = sizeof(struct rga_vb_buffer);
 	dst_vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
 	dst_vq->lock = &ctx->rga->mutex;
 	dst_vq->dev = ctx->rga->v4l2_dev.dev;
@@ -872,26 +869,13 @@ static int rga_probe(struct platform_device *pdev)
 		goto rel_m2m;
 	}
 
-	rga->src_mmu_pages =
-		(struct rga_dma_desc *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 3);
-	if (!rga->src_mmu_pages) {
-		ret = -ENOMEM;
-		goto free_dma;
-	}
-	rga->dst_mmu_pages =
-		(struct rga_dma_desc *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 3);
-	if (!rga->dst_mmu_pages) {
-		ret = -ENOMEM;
-		goto free_src_pages;
-	}
-
 	def_frame.stride = (def_frame.width * def_frame.fmt->depth) >> 3;
 	def_frame.size = def_frame.stride * def_frame.height;
 
 	ret = video_register_device(vfd, VFL_TYPE_VIDEO, -1);
 	if (ret) {
 		v4l2_err(&rga->v4l2_dev, "Failed to register video device\n");
-		goto free_dst_pages;
+		goto free_dma;
 	}
 
 	v4l2_info(&rga->v4l2_dev, "Registered %s as /dev/%s\n",
@@ -899,10 +883,6 @@ static int rga_probe(struct platform_device *pdev)
 
 	return 0;
 
-free_dst_pages:
-	free_pages((unsigned long)rga->dst_mmu_pages, 3);
-free_src_pages:
-	free_pages((unsigned long)rga->src_mmu_pages, 3);
 free_dma:
 	dma_free_attrs(rga->dev, RGA_CMDBUF_SIZE, rga->cmdbuf_virt,
 		       rga->cmdbuf_phy, DMA_ATTR_WRITE_COMBINE);
@@ -925,9 +905,6 @@ static void rga_remove(struct platform_device *pdev)
 	dma_free_attrs(rga->dev, RGA_CMDBUF_SIZE, rga->cmdbuf_virt,
 		       rga->cmdbuf_phy, DMA_ATTR_WRITE_COMBINE);
 
-	free_pages((unsigned long)rga->src_mmu_pages, 3);
-	free_pages((unsigned long)rga->dst_mmu_pages, 3);
-
 	v4l2_info(&rga->v4l2_dev, "Removing\n");
 
 	v4l2_m2m_release(rga->m2m_dev);
diff --git a/drivers/media/platform/rockchip/rga/rga.h b/drivers/media/platform/rockchip/rga/rga.h
index 22f7da28ac51..678f0f627bdd 100644
--- a/drivers/media/platform/rockchip/rga/rga.h
+++ b/drivers/media/platform/rockchip/rga/rga.h
@@ -85,15 +85,27 @@ struct rockchip_rga {
 	struct rga_ctx *curr;
 	dma_addr_t cmdbuf_phy;
 	void *cmdbuf_virt;
-	struct rga_dma_desc *src_mmu_pages;
-	struct rga_dma_desc *dst_mmu_pages;
 };
 
+struct rga_vb_buffer {
+	struct vb2_v4l2_buffer vb_buf;
+	struct list_head queue;
+
+	/* RGA MMU mapping for this buffer */
+	struct rga_dma_desc *dma_desc;
+	dma_addr_t dma_desc_pa;
+	size_t n_desc;
+};
+
+static inline struct rga_vb_buffer *vb_to_rga(struct vb2_v4l2_buffer *vb)
+{
+	return container_of(vb, struct rga_vb_buffer, vb_buf);
+}
+
 struct rga_frame *rga_get_frame(struct rga_ctx *ctx, enum v4l2_buf_type type);
 
 /* RGA Buffers Manage */
 extern const struct vb2_ops rga_qops;
-void rga_buf_map(struct vb2_buffer *vb);
 
 /* RGA Hardware */
 static inline void rga_write(struct rockchip_rga *rga, u32 reg, u32 value)
@@ -114,6 +126,7 @@ static inline void rga_mod(struct rockchip_rga *rga, u32 reg, u32 val, u32 mask)
 	rga_write(rga, reg, temp);
 };
 
-void rga_hw_start(struct rockchip_rga *rga);
+void rga_hw_start(struct rockchip_rga *rga,
+		  struct rga_vb_buffer *src, struct rga_vb_buffer *dst);
 
 #endif

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 04/13] media: rockchip: rga: split src and dst buffer setup

[Michael Tretter]

Split the register setup for the source and destination video buffers
into separate functions.

This is a cleanup to make the code more readable by separating the
offset calculation for the different buffers and prepares the driver for
using pre-calculated offsets of planes.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga-hw.c | 94 ++++++++++++++++++++--------
 1 file changed, 68 insertions(+), 26 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga-hw.c b/drivers/media/platform/rockchip/rga/rga-hw.c
index b391d97d4632..836ec7721b21 100644
--- a/drivers/media/platform/rockchip/rga/rga-hw.c
+++ b/drivers/media/platform/rockchip/rga/rga-hw.c
@@ -163,7 +163,7 @@ static void rga_cmd_set_trans_info(struct rga_ctx *ctx)
 	struct rockchip_rga *rga = ctx->rga;
 	u32 *dest = rga->cmdbuf_virt;
 	unsigned int scale_dst_w, scale_dst_h;
-	unsigned int src_h, src_w, src_x, src_y, dst_h, dst_w, dst_x, dst_y;
+	unsigned int src_h, src_w, dst_h, dst_w;
 	union rga_src_info src_info;
 	union rga_dst_info dst_info;
 	union rga_src_x_factor x_factor;
@@ -173,18 +173,10 @@ static void rga_cmd_set_trans_info(struct rga_ctx *ctx)
 	union rga_dst_vir_info dst_vir_info;
 	union rga_dst_act_info dst_act_info;
 
-	struct rga_addr_offset *dst_offset;
-	struct rga_corners_addr_offset offsets;
-	struct rga_corners_addr_offset src_offsets;
-
 	src_h = ctx->in.crop.height;
 	src_w = ctx->in.crop.width;
-	src_x = ctx->in.crop.left;
-	src_y = ctx->in.crop.top;
 	dst_h = ctx->out.crop.height;
 	dst_w = ctx->out.crop.width;
-	dst_x = ctx->out.crop.left;
-	dst_y = ctx->out.crop.top;
 
 	src_info.val = dest[(RGA_SRC_INFO - RGA_MODE_BASE_REG) >> 2];
 	dst_info.val = dest[(RGA_DST_INFO - RGA_MODE_BASE_REG) >> 2];
@@ -312,32 +304,85 @@ static void rga_cmd_set_trans_info(struct rga_ctx *ctx)
 	dst_act_info.data.act_height = dst_h - 1;
 	dst_act_info.data.act_width = dst_w - 1;
 
+	dest[(RGA_SRC_X_FACTOR - RGA_MODE_BASE_REG) >> 2] = x_factor.val;
+	dest[(RGA_SRC_Y_FACTOR - RGA_MODE_BASE_REG) >> 2] = y_factor.val;
+	dest[(RGA_SRC_VIR_INFO - RGA_MODE_BASE_REG) >> 2] = src_vir_info.val;
+	dest[(RGA_SRC_ACT_INFO - RGA_MODE_BASE_REG) >> 2] = src_act_info.val;
+
+	dest[(RGA_SRC_INFO - RGA_MODE_BASE_REG) >> 2] = src_info.val;
+
+	dest[(RGA_DST_VIR_INFO - RGA_MODE_BASE_REG) >> 2] = dst_vir_info.val;
+	dest[(RGA_DST_ACT_INFO - RGA_MODE_BASE_REG) >> 2] = dst_act_info.val;
+
+	dest[(RGA_DST_INFO - RGA_MODE_BASE_REG) >> 2] = dst_info.val;
+}
+
+static void rga_cmd_set_src_info(struct rga_ctx *ctx)
+{
+	struct rga_corners_addr_offset src_offsets;
+	struct rockchip_rga *rga = ctx->rga;
+	u32 *dest = rga->cmdbuf_virt;
+	unsigned int src_h, src_w, src_x, src_y;
+
+	src_h = ctx->in.crop.height;
+	src_w = ctx->in.crop.width;
+	src_x = ctx->in.crop.left;
+	src_y = ctx->in.crop.top;
+
 	/*
 	 * Calculate the source framebuffer base address with offset pixel.
 	 */
 	src_offsets = rga_get_addr_offset(&ctx->in, src_x, src_y,
 					  src_w, src_h);
 
-	/*
-	 * Configure the dest framebuffer base address with pixel offset.
-	 */
-	offsets = rga_get_addr_offset(&ctx->out, dst_x, dst_y, dst_w, dst_h);
-	dst_offset = rga_lookup_draw_pos(&offsets, src_info.data.rot_mode,
-					 src_info.data.mir_mode);
-
 	dest[(RGA_SRC_Y_RGB_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
 		src_offsets.left_top.y_off;
 	dest[(RGA_SRC_CB_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
 		src_offsets.left_top.u_off;
 	dest[(RGA_SRC_CR_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
 		src_offsets.left_top.v_off;
+}
 
-	dest[(RGA_SRC_X_FACTOR - RGA_MODE_BASE_REG) >> 2] = x_factor.val;
-	dest[(RGA_SRC_Y_FACTOR - RGA_MODE_BASE_REG) >> 2] = y_factor.val;
-	dest[(RGA_SRC_VIR_INFO - RGA_MODE_BASE_REG) >> 2] = src_vir_info.val;
-	dest[(RGA_SRC_ACT_INFO - RGA_MODE_BASE_REG) >> 2] = src_act_info.val;
+static void rga_cmd_set_dst_info(struct rga_ctx *ctx)
+{
+	struct rga_addr_offset *dst_offset;
+	struct rga_corners_addr_offset offsets;
+	struct rockchip_rga *rga = ctx->rga;
+	u32 *dest = rga->cmdbuf_virt;
+	unsigned int dst_h, dst_w, dst_x, dst_y;
+	unsigned int mir_mode = 0;
+	unsigned int rot_mode = 0;
 
-	dest[(RGA_SRC_INFO - RGA_MODE_BASE_REG) >> 2] = src_info.val;
+	dst_h = ctx->out.crop.height;
+	dst_w = ctx->out.crop.width;
+	dst_x = ctx->out.crop.left;
+	dst_y = ctx->out.crop.top;
+
+	if (ctx->vflip)
+		mir_mode |= RGA_SRC_MIRR_MODE_X;
+	if (ctx->hflip)
+		mir_mode |= RGA_SRC_MIRR_MODE_Y;
+
+	switch (ctx->rotate) {
+	case 90:
+		rot_mode = RGA_SRC_ROT_MODE_90_DEGREE;
+		break;
+	case 180:
+		rot_mode = RGA_SRC_ROT_MODE_180_DEGREE;
+		break;
+	case 270:
+		rot_mode = RGA_SRC_ROT_MODE_270_DEGREE;
+		break;
+	default:
+		rot_mode = RGA_SRC_ROT_MODE_0_DEGREE;
+		break;
+	}
+
+	/*
+	 * Configure the dest framebuffer base address with pixel offset.
+	 */
+	offsets = rga_get_addr_offset(&ctx->out, dst_x, dst_y, dst_w, dst_h);
+	dst_offset = rga_lookup_draw_pos(&offsets, mir_mode, rot_mode);
 
 	dest[(RGA_DST_Y_RGB_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
 		dst_offset->y_off;
@@ -345,11 +390,6 @@ static void rga_cmd_set_trans_info(struct rga_ctx *ctx)
 		dst_offset->u_off;
 	dest[(RGA_DST_CR_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
 		dst_offset->v_off;
-
-	dest[(RGA_DST_VIR_INFO - RGA_MODE_BASE_REG) >> 2] = dst_vir_info.val;
-	dest[(RGA_DST_ACT_INFO - RGA_MODE_BASE_REG) >> 2] = dst_act_info.val;
-
-	dest[(RGA_DST_INFO - RGA_MODE_BASE_REG) >> 2] = dst_info.val;
 }
 
 static void rga_cmd_set_mode(struct rga_ctx *ctx)
@@ -392,6 +432,8 @@ static void rga_cmd_set(struct rga_ctx *ctx,
 	rga_cmd_set_dst_addr(ctx, dst->dma_desc_pa);
 	rga_cmd_set_mode(ctx);
 
+	rga_cmd_set_src_info(ctx);
+	rga_cmd_set_dst_info(ctx);
 	rga_cmd_set_trans_info(ctx);
 
 	rga_write(rga, RGA_CMD_BASE, rga->cmdbuf_phy);

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 05/13] media: rockchip: rga: pre-calculate plane offsets

[Michael Tretter]

Calculate the plane offsets and store them with the video buffer while
creating the buffer mapping.

This allows the driver to more freely handle the memory of the DMA
mapping as the offsets and the mapping can be kept in sync.

The driver still has to update the offsets to respect the configured
cropping and rotation, but this calculation is now separated from the
calculation of the plane offsets.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2:
- Drop unused uv_factor to fix smatch warning
---
 drivers/media/platform/rockchip/rga/rga-buf.c | 16 ++++++++++++
 drivers/media/platform/rockchip/rga/rga-hw.c  | 36 +++++++++++----------------
 drivers/media/platform/rockchip/rga/rga.h     |  9 +++++++
 3 files changed, 40 insertions(+), 21 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga-buf.c b/drivers/media/platform/rockchip/rga/rga-buf.c
index ff36efade060..d6774a844d9b 100644
--- a/drivers/media/platform/rockchip/rga/rga-buf.c
+++ b/drivers/media/platform/rockchip/rga/rga-buf.c
@@ -74,6 +74,18 @@ static int rga_buf_init(struct vb2_buffer *vb)
 	return 0;
 }
 
+static int get_plane_offset(struct rga_frame *f, int plane)
+{
+	if (plane == 0)
+		return 0;
+	if (plane == 1)
+		return f->width * f->height;
+	if (plane == 2)
+		return f->width * f->height + (f->width * f->height / f->fmt->uv_factor);
+
+	return -EINVAL;
+}
+
 static int rga_buf_prepare(struct vb2_buffer *vb)
 {
 	struct vb2_v4l2_buffer *vbuf = to_vb2_v4l2_buffer(vb);
@@ -89,6 +101,10 @@ static int rga_buf_prepare(struct vb2_buffer *vb)
 	/* Create local MMU table for RGA */
 	fill_descriptors(rbuf->dma_desc, vb2_dma_sg_plane_desc(vb, 0));
 
+	rbuf->offset.y_off = get_plane_offset(f, 0);
+	rbuf->offset.u_off = get_plane_offset(f, 1);
+	rbuf->offset.v_off = get_plane_offset(f, 2);
+
 	return 0;
 }
 
diff --git a/drivers/media/platform/rockchip/rga/rga-hw.c b/drivers/media/platform/rockchip/rga/rga-hw.c
index 836ec7721b21..11c3d7234757 100644
--- a/drivers/media/platform/rockchip/rga/rga-hw.c
+++ b/drivers/media/platform/rockchip/rga/rga-hw.c
@@ -16,12 +16,6 @@ enum e_rga_start_pos {
 	RB = 3,
 };
 
-struct rga_addr_offset {
-	unsigned int y_off;
-	unsigned int u_off;
-	unsigned int v_off;
-};
-
 struct rga_corners_addr_offset {
 	struct rga_addr_offset left_top;
 	struct rga_addr_offset right_top;
@@ -43,13 +37,13 @@ static unsigned int rga_get_scaling(unsigned int src, unsigned int dst)
 }
 
 static struct rga_corners_addr_offset
-rga_get_addr_offset(struct rga_frame *frm, unsigned int x, unsigned int y,
-		    unsigned int w, unsigned int h)
+rga_get_addr_offset(struct rga_frame *frm, struct rga_addr_offset *offset,
+		    unsigned int x, unsigned int y, unsigned int w, unsigned int h)
 {
 	struct rga_corners_addr_offset offsets;
 	struct rga_addr_offset *lt, *lb, *rt, *rb;
 	unsigned int x_div = 0,
-		     y_div = 0, uv_stride = 0, pixel_width = 0, uv_factor = 0;
+		     y_div = 0, uv_stride = 0, pixel_width = 0;
 
 	lt = &offsets.left_top;
 	lb = &offsets.left_bottom;
@@ -58,14 +52,12 @@ rga_get_addr_offset(struct rga_frame *frm, unsigned int x, unsigned int y,
 
 	x_div = frm->fmt->x_div;
 	y_div = frm->fmt->y_div;
-	uv_factor = frm->fmt->uv_factor;
 	uv_stride = frm->stride / x_div;
 	pixel_width = frm->stride / frm->width;
 
-	lt->y_off = y * frm->stride + x * pixel_width;
-	lt->u_off =
-		frm->width * frm->height + (y / y_div) * uv_stride + x / x_div;
-	lt->v_off = lt->u_off + frm->width * frm->height / uv_factor;
+	lt->y_off = offset->y_off + y * frm->stride + x * pixel_width;
+	lt->u_off = offset->u_off + (y / y_div) * uv_stride + x / x_div;
+	lt->v_off = offset->v_off + (y / y_div) * uv_stride + x / x_div;
 
 	lb->y_off = lt->y_off + (h - 1) * frm->stride;
 	lb->u_off = lt->u_off + (h / y_div - 1) * uv_stride;
@@ -317,7 +309,8 @@ static void rga_cmd_set_trans_info(struct rga_ctx *ctx)
 	dest[(RGA_DST_INFO - RGA_MODE_BASE_REG) >> 2] = dst_info.val;
 }
 
-static void rga_cmd_set_src_info(struct rga_ctx *ctx)
+static void rga_cmd_set_src_info(struct rga_ctx *ctx,
+				 struct rga_addr_offset *offset)
 {
 	struct rga_corners_addr_offset src_offsets;
 	struct rockchip_rga *rga = ctx->rga;
@@ -332,8 +325,8 @@ static void rga_cmd_set_src_info(struct rga_ctx *ctx)
 	/*
 	 * Calculate the source framebuffer base address with offset pixel.
 	 */
-	src_offsets = rga_get_addr_offset(&ctx->in, src_x, src_y,
-					  src_w, src_h);
+	src_offsets = rga_get_addr_offset(&ctx->in, offset,
+					  src_x, src_y, src_w, src_h);
 
 	dest[(RGA_SRC_Y_RGB_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
 		src_offsets.left_top.y_off;
@@ -343,7 +336,8 @@ static void rga_cmd_set_src_info(struct rga_ctx *ctx)
 		src_offsets.left_top.v_off;
 }
 
-static void rga_cmd_set_dst_info(struct rga_ctx *ctx)
+static void rga_cmd_set_dst_info(struct rga_ctx *ctx,
+				 struct rga_addr_offset *offset)
 {
 	struct rga_addr_offset *dst_offset;
 	struct rga_corners_addr_offset offsets;
@@ -381,7 +375,7 @@ static void rga_cmd_set_dst_info(struct rga_ctx *ctx)
 	/*
 	 * Configure the dest framebuffer base address with pixel offset.
 	 */
-	offsets = rga_get_addr_offset(&ctx->out, dst_x, dst_y, dst_w, dst_h);
+	offsets = rga_get_addr_offset(&ctx->out, offset, dst_x, dst_y, dst_w, dst_h);
 	dst_offset = rga_lookup_draw_pos(&offsets, mir_mode, rot_mode);
 
 	dest[(RGA_DST_Y_RGB_BASE_ADDR - RGA_MODE_BASE_REG) >> 2] =
@@ -432,8 +426,8 @@ static void rga_cmd_set(struct rga_ctx *ctx,
 	rga_cmd_set_dst_addr(ctx, dst->dma_desc_pa);
 	rga_cmd_set_mode(ctx);
 
-	rga_cmd_set_src_info(ctx);
-	rga_cmd_set_dst_info(ctx);
+	rga_cmd_set_src_info(ctx, &src->offset);
+	rga_cmd_set_dst_info(ctx, &dst->offset);
 	rga_cmd_set_trans_info(ctx);
 
 	rga_write(rga, RGA_CMD_BASE, rga->cmdbuf_phy);
diff --git a/drivers/media/platform/rockchip/rga/rga.h b/drivers/media/platform/rockchip/rga/rga.h
index 678f0f627bdd..f3953a07d9e7 100644
--- a/drivers/media/platform/rockchip/rga/rga.h
+++ b/drivers/media/platform/rockchip/rga/rga.h
@@ -87,6 +87,12 @@ struct rockchip_rga {
 	void *cmdbuf_virt;
 };
 
+struct rga_addr_offset {
+	unsigned int y_off;
+	unsigned int u_off;
+	unsigned int v_off;
+};
+
 struct rga_vb_buffer {
 	struct vb2_v4l2_buffer vb_buf;
 	struct list_head queue;
@@ -95,6 +101,9 @@ struct rga_vb_buffer {
 	struct rga_dma_desc *dma_desc;
 	dma_addr_t dma_desc_pa;
 	size_t n_desc;
+
+	/* Plane offsets of this buffer into the mapping */
+	struct rga_addr_offset offset;
 };
 
 static inline struct rga_vb_buffer *vb_to_rga(struct vb2_v4l2_buffer *vb)

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 06/13] media: rockchip: rga: set dma mask to 32 bits

[Michael Tretter]

The RGA DMA descriptor list contains only 32-bit addresses. Set the
dma_mask to only allocate memory that is addressable by the descriptors.

This prevents errors when preparing vb2 buffers that were allocated by
the RGA.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2:
- Add mask for DMA coherent
- Set gfp_flags to DMA32
---
 drivers/media/platform/rockchip/rga/rga.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index f18fccc7b204..ea566c11734a 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -98,6 +98,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
 	src_vq->drv_priv = ctx;
 	src_vq->ops = &rga_qops;
 	src_vq->mem_ops = &vb2_dma_sg_memops;
+	dst_vq->gfp_flags = __GFP_DMA32;
 	src_vq->buf_struct_size = sizeof(struct rga_vb_buffer);
 	src_vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
 	src_vq->lock = &ctx->rga->mutex;
@@ -112,6 +113,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
 	dst_vq->drv_priv = ctx;
 	dst_vq->ops = &rga_qops;
 	dst_vq->mem_ops = &vb2_dma_sg_memops;
+	dst_vq->gfp_flags = __GFP_DMA32;
 	dst_vq->buf_struct_size = sizeof(struct rga_vb_buffer);
 	dst_vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
 	dst_vq->lock = &ctx->rga->mutex;
@@ -824,6 +826,12 @@ static int rga_probe(struct platform_device *pdev)
 		goto err_put_clk;
 	}
 
+	ret = dma_set_mask_and_coherent(rga->dev, DMA_BIT_MASK(32));
+	if (ret) {
+		dev_err(rga->dev, "32-bit DMA not supported");
+		goto err_put_clk;
+	}
+
 	ret = v4l2_device_register(&pdev->dev, &rga->v4l2_dev);
 	if (ret)
 		goto err_put_clk;

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 07/13] media: rockchip: rga: use clamp() to clamp size to limits

[Michael Tretter]

The try_fmt should limit the width and height to the know limits of the
RGA. Use the clamp() helper instead of open coding the clamping.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 13 ++++---------
 1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index ea566c11734a..792084918798 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -482,15 +482,10 @@ static int vidioc_try_fmt(struct file *file, void *prv, struct v4l2_format *f)
 
 	f->fmt.pix.field = V4L2_FIELD_NONE;
 
-	if (f->fmt.pix.width > MAX_WIDTH)
-		f->fmt.pix.width = MAX_WIDTH;
-	if (f->fmt.pix.height > MAX_HEIGHT)
-		f->fmt.pix.height = MAX_HEIGHT;
-
-	if (f->fmt.pix.width < MIN_WIDTH)
-		f->fmt.pix.width = MIN_WIDTH;
-	if (f->fmt.pix.height < MIN_HEIGHT)
-		f->fmt.pix.height = MIN_HEIGHT;
+	f->fmt.pix.width = clamp(f->fmt.pix.width,
+				 (u32)MIN_WIDTH, (u32)MAX_WIDTH);
+	f->fmt.pix.height = clamp(f->fmt.pix.height,
+				  (u32)MIN_HEIGHT, (u32)MAX_HEIGHT);
 
 	if (fmt->hw_format >= RGA_COLOR_FMT_YUV422SP)
 		f->fmt.pix.bytesperline = f->fmt.pix.width;

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 08/13] media: rockchip: rga: use pixelformat to find format

[Michael Tretter]

Use the pixelformat instead of the v4l2_format to find the rga_fmt. This
avoids knowing the structure and type of v4l2_format in rga_fmt_find and
simplifies the function.

Also cleanup the users of the function. In try_fmt always return the
found pixel format to make sure that the pixel format is always set.
Thus, we can be sure that we will find the rga_fmt in s_fmt and can drop
the check if a given format has been found.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 18 +++++++-----------
 1 file changed, 7 insertions(+), 11 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index 792084918798..8ae68890a8f7 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -320,12 +320,12 @@ static struct rga_fmt formats[] = {
 
 #define NUM_FORMATS ARRAY_SIZE(formats)
 
-static struct rga_fmt *rga_fmt_find(struct v4l2_format *f)
+static struct rga_fmt *rga_fmt_find(u32 pixelformat)
 {
 	unsigned int i;
 
 	for (i = 0; i < NUM_FORMATS; i++) {
-		if (formats[i].fourcc == f->fmt.pix.pixelformat)
+		if (formats[i].fourcc == pixelformat)
 			return &formats[i];
 	}
 	return NULL;
@@ -474,11 +474,11 @@ static int vidioc_try_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
 	struct rga_fmt *fmt;
 
-	fmt = rga_fmt_find(f);
-	if (!fmt) {
+	fmt = rga_fmt_find(f->fmt.pix.pixelformat);
+	if (!fmt)
 		fmt = &formats[0];
-		f->fmt.pix.pixelformat = fmt->fourcc;
-	}
+
+	f->fmt.pix.pixelformat = fmt->fourcc;
 
 	f->fmt.pix.field = V4L2_FIELD_NONE;
 
@@ -504,7 +504,6 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	struct rockchip_rga *rga = ctx->rga;
 	struct vb2_queue *vq;
 	struct rga_frame *frm;
-	struct rga_fmt *fmt;
 	int ret = 0;
 
 	/* Adjust all values accordingly to the hardware capabilities
@@ -521,13 +520,10 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	frm = rga_get_frame(ctx, f->type);
 	if (IS_ERR(frm))
 		return PTR_ERR(frm);
-	fmt = rga_fmt_find(f);
-	if (!fmt)
-		return -EINVAL;
 	frm->width = f->fmt.pix.width;
 	frm->height = f->fmt.pix.height;
 	frm->size = f->fmt.pix.sizeimage;
-	frm->fmt = fmt;
+	frm->fmt = rga_fmt_find(f->fmt.pix.pixelformat);
 	frm->stride = f->fmt.pix.bytesperline;
 	frm->colorspace = f->fmt.pix.colorspace;
 

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 09/13] media: rockchip: rga: add local variable for pix_format

[Michael Tretter]

The local variable allows to simplify the accessed to the format and
makes it easier to change the type of the format.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 55 ++++++++++++++++++-------------
 1 file changed, 32 insertions(+), 23 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index 8ae68890a8f7..f5d0d87cce22 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -448,6 +448,7 @@ static int vidioc_enum_fmt(struct file *file, void *prv, struct v4l2_fmtdesc *f)
 
 static int vidioc_g_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
+	struct v4l2_pix_format *pix_fmt = &f->fmt.pix;
 	struct rga_ctx *ctx = prv;
 	struct vb2_queue *vq;
 	struct rga_frame *frm;
@@ -459,47 +460,49 @@ static int vidioc_g_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	if (IS_ERR(frm))
 		return PTR_ERR(frm);
 
-	f->fmt.pix.width = frm->width;
-	f->fmt.pix.height = frm->height;
-	f->fmt.pix.field = V4L2_FIELD_NONE;
-	f->fmt.pix.pixelformat = frm->fmt->fourcc;
-	f->fmt.pix.bytesperline = frm->stride;
-	f->fmt.pix.sizeimage = frm->size;
-	f->fmt.pix.colorspace = frm->colorspace;
+	pix_fmt->width = frm->width;
+	pix_fmt->height = frm->height;
+	pix_fmt->field = V4L2_FIELD_NONE;
+	pix_fmt->pixelformat = frm->fmt->fourcc;
+	pix_fmt->bytesperline = frm->stride;
+	pix_fmt->sizeimage = frm->size;
+	pix_fmt->colorspace = frm->colorspace;
 
 	return 0;
 }
 
 static int vidioc_try_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
+	struct v4l2_pix_format *pix_fmt = &f->fmt.pix;
 	struct rga_fmt *fmt;
 
 	fmt = rga_fmt_find(f->fmt.pix.pixelformat);
 	if (!fmt)
 		fmt = &formats[0];
 
-	f->fmt.pix.pixelformat = fmt->fourcc;
+	pix_fmt->pixelformat = fmt->fourcc;
 
-	f->fmt.pix.field = V4L2_FIELD_NONE;
+	pix_fmt->field = V4L2_FIELD_NONE;
 
-	f->fmt.pix.width = clamp(f->fmt.pix.width,
-				 (u32)MIN_WIDTH, (u32)MAX_WIDTH);
-	f->fmt.pix.height = clamp(f->fmt.pix.height,
-				  (u32)MIN_HEIGHT, (u32)MAX_HEIGHT);
+	pix_fmt->width = clamp(pix_fmt->width,
+			       (u32)MIN_WIDTH, (u32)MAX_WIDTH);
+	pix_fmt->height = clamp(pix_fmt->height,
+				(u32)MIN_HEIGHT, (u32)MAX_HEIGHT);
 
 	if (fmt->hw_format >= RGA_COLOR_FMT_YUV422SP)
-		f->fmt.pix.bytesperline = f->fmt.pix.width;
+		pix_fmt->bytesperline = pix_fmt->width;
 	else
-		f->fmt.pix.bytesperline = (f->fmt.pix.width * fmt->depth) >> 3;
+		pix_fmt->bytesperline = (pix_fmt->width * fmt->depth) >> 3;
 
-	f->fmt.pix.sizeimage =
-		f->fmt.pix.height * (f->fmt.pix.width * fmt->depth) >> 3;
+	pix_fmt->sizeimage =
+		pix_fmt->height * (pix_fmt->width * fmt->depth) >> 3;
 
 	return 0;
 }
 
 static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
+	struct v4l2_pix_format *pix_fmt = &f->fmt.pix;
 	struct rga_ctx *ctx = prv;
 	struct rockchip_rga *rga = ctx->rga;
 	struct vb2_queue *vq;
@@ -520,12 +523,12 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	frm = rga_get_frame(ctx, f->type);
 	if (IS_ERR(frm))
 		return PTR_ERR(frm);
-	frm->width = f->fmt.pix.width;
-	frm->height = f->fmt.pix.height;
-	frm->size = f->fmt.pix.sizeimage;
-	frm->fmt = rga_fmt_find(f->fmt.pix.pixelformat);
-	frm->stride = f->fmt.pix.bytesperline;
-	frm->colorspace = f->fmt.pix.colorspace;
+	frm->width = pix_fmt->width;
+	frm->height = pix_fmt->height;
+	frm->size = pix_fmt->sizeimage;
+	frm->fmt = rga_fmt_find(pix_fmt->pixelformat);
+	frm->stride = pix_fmt->bytesperline;
+	frm->colorspace = pix_fmt->colorspace;
 
 	/* Reset crop settings */
 	frm->crop.left = 0;
@@ -533,6 +536,12 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	frm->crop.width = frm->width;
 	frm->crop.height = frm->height;
 
+	v4l2_dbg(debug, 1, &rga->v4l2_dev,
+		 "[%s] fmt - %p4cc %dx%d (stride %d, sizeimage %d)\n",
+		  V4L2_TYPE_IS_OUTPUT(f->type) ? "OUTPUT" : "CAPTURE",
+		  &frm->fmt->fourcc, frm->width, frm->height,
+		  frm->stride, frm->size);
+
 	return 0;
 }
 

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 10/13] media: rockchip: rga: use macros for testing buffer type

[Michael Tretter]

Use the provided V4L2_TYPE_IS_{OUTPUT,CAPTURE} macros to check if the
buffer or queue is OUTPUT or CAPTURE. The macros work also work for the
_MPLANE buffer and queue types and make it easier to switch to the
multi-planar API.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 21 +++++++++------------
 1 file changed, 9 insertions(+), 12 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index f5d0d87cce22..91be87e01dab 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -344,14 +344,11 @@ static struct rga_frame def_frame = {
 
 struct rga_frame *rga_get_frame(struct rga_ctx *ctx, enum v4l2_buf_type type)
 {
-	switch (type) {
-	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
+	if (V4L2_TYPE_IS_OUTPUT(type))
 		return &ctx->in;
-	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
+	if (V4L2_TYPE_IS_CAPTURE(type))
 		return &ctx->out;
-	default:
-		return ERR_PTR(-EINVAL);
-	}
+	return ERR_PTR(-EINVAL);
 }
 
 static int rga_open(struct file *file)
@@ -559,21 +556,21 @@ static int vidioc_g_selection(struct file *file, void *prv,
 	switch (s->target) {
 	case V4L2_SEL_TGT_COMPOSE_DEFAULT:
 	case V4L2_SEL_TGT_COMPOSE_BOUNDS:
-		if (s->type != V4L2_BUF_TYPE_VIDEO_CAPTURE)
+		if (!V4L2_TYPE_IS_CAPTURE(s->type))
 			return -EINVAL;
 		break;
 	case V4L2_SEL_TGT_CROP_DEFAULT:
 	case V4L2_SEL_TGT_CROP_BOUNDS:
-		if (s->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		if (!V4L2_TYPE_IS_OUTPUT(s->type))
 			return -EINVAL;
 		break;
 	case V4L2_SEL_TGT_COMPOSE:
-		if (s->type != V4L2_BUF_TYPE_VIDEO_CAPTURE)
+		if (!V4L2_TYPE_IS_CAPTURE(s->type))
 			return -EINVAL;
 		use_frame = true;
 		break;
 	case V4L2_SEL_TGT_CROP:
-		if (s->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		if (!V4L2_TYPE_IS_OUTPUT(s->type))
 			return -EINVAL;
 		use_frame = true;
 		break;
@@ -611,7 +608,7 @@ static int vidioc_s_selection(struct file *file, void *prv,
 		 * COMPOSE target is only valid for capture buffer type, return
 		 * error for output buffer type
 		 */
-		if (s->type != V4L2_BUF_TYPE_VIDEO_CAPTURE)
+		if (!V4L2_TYPE_IS_CAPTURE(s->type))
 			return -EINVAL;
 		break;
 	case V4L2_SEL_TGT_CROP:
@@ -619,7 +616,7 @@ static int vidioc_s_selection(struct file *file, void *prv,
 		 * CROP target is only valid for output buffer type, return
 		 * error for capture buffer type
 		 */
-		if (s->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		if (!V4L2_TYPE_IS_OUTPUT(s->type))
 			return -EINVAL;
 		break;
 	/*

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 11/13] media: rockchip: rga: switch to multi-planar API

[Michael Tretter]

Switch to the multi-planar API, which allows to handle buffers with
separate planes.

The RGA driver doesn't expose multi-planar formats, yet. The existing
contiguous planar formats can be used with the multi-planar API as well,
but the multi-planar API is required for multi-planar formats.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 60 +++++++++++++++----------------
 1 file changed, 29 insertions(+), 31 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index 91be87e01dab..f549966c3649 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -93,7 +93,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
 	struct rga_ctx *ctx = priv;
 	int ret;
 
-	src_vq->type = V4L2_BUF_TYPE_VIDEO_OUTPUT;
+	src_vq->type = V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE;
 	src_vq->io_modes = VB2_MMAP | VB2_DMABUF;
 	src_vq->drv_priv = ctx;
 	src_vq->ops = &rga_qops;
@@ -108,7 +108,7 @@ queue_init(void *priv, struct vb2_queue *src_vq, struct vb2_queue *dst_vq)
 	if (ret)
 		return ret;
 
-	dst_vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
+	dst_vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE;
 	dst_vq->io_modes = VB2_MMAP | VB2_DMABUF;
 	dst_vq->drv_priv = ctx;
 	dst_vq->ops = &rga_qops;
@@ -445,7 +445,7 @@ static int vidioc_enum_fmt(struct file *file, void *prv, struct v4l2_fmtdesc *f)
 
 static int vidioc_g_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
-	struct v4l2_pix_format *pix_fmt = &f->fmt.pix;
+	struct v4l2_pix_format_mplane *pix_fmt = &f->fmt.pix_mp;
 	struct rga_ctx *ctx = prv;
 	struct vb2_queue *vq;
 	struct rga_frame *frm;
@@ -457,12 +457,9 @@ static int vidioc_g_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	if (IS_ERR(frm))
 		return PTR_ERR(frm);
 
-	pix_fmt->width = frm->width;
-	pix_fmt->height = frm->height;
+	v4l2_fill_pixfmt_mp(pix_fmt, frm->fmt->fourcc, frm->width, frm->height);
+
 	pix_fmt->field = V4L2_FIELD_NONE;
-	pix_fmt->pixelformat = frm->fmt->fourcc;
-	pix_fmt->bytesperline = frm->stride;
-	pix_fmt->sizeimage = frm->size;
 	pix_fmt->colorspace = frm->colorspace;
 
 	return 0;
@@ -470,41 +467,33 @@ static int vidioc_g_fmt(struct file *file, void *prv, struct v4l2_format *f)
 
 static int vidioc_try_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
-	struct v4l2_pix_format *pix_fmt = &f->fmt.pix;
+	struct v4l2_pix_format_mplane *pix_fmt = &f->fmt.pix_mp;
 	struct rga_fmt *fmt;
 
-	fmt = rga_fmt_find(f->fmt.pix.pixelformat);
+	fmt = rga_fmt_find(pix_fmt->pixelformat);
 	if (!fmt)
 		fmt = &formats[0];
 
-	pix_fmt->pixelformat = fmt->fourcc;
-
-	pix_fmt->field = V4L2_FIELD_NONE;
-
 	pix_fmt->width = clamp(pix_fmt->width,
 			       (u32)MIN_WIDTH, (u32)MAX_WIDTH);
 	pix_fmt->height = clamp(pix_fmt->height,
 				(u32)MIN_HEIGHT, (u32)MAX_HEIGHT);
 
-	if (fmt->hw_format >= RGA_COLOR_FMT_YUV422SP)
-		pix_fmt->bytesperline = pix_fmt->width;
-	else
-		pix_fmt->bytesperline = (pix_fmt->width * fmt->depth) >> 3;
-
-	pix_fmt->sizeimage =
-		pix_fmt->height * (pix_fmt->width * fmt->depth) >> 3;
+	v4l2_fill_pixfmt_mp(pix_fmt, fmt->fourcc, pix_fmt->width, pix_fmt->height);
+	pix_fmt->field = V4L2_FIELD_NONE;
 
 	return 0;
 }
 
 static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 {
-	struct v4l2_pix_format *pix_fmt = &f->fmt.pix;
+	struct v4l2_pix_format_mplane *pix_fmt = &f->fmt.pix_mp;
 	struct rga_ctx *ctx = prv;
 	struct rockchip_rga *rga = ctx->rga;
 	struct vb2_queue *vq;
 	struct rga_frame *frm;
 	int ret = 0;
+	int i;
 
 	/* Adjust all values accordingly to the hardware capabilities
 	 * and chosen format.
@@ -522,9 +511,11 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 		return PTR_ERR(frm);
 	frm->width = pix_fmt->width;
 	frm->height = pix_fmt->height;
-	frm->size = pix_fmt->sizeimage;
+	frm->size = 0;
+	for (i = 0; i < pix_fmt->num_planes; i++)
+		frm->size += pix_fmt->plane_fmt[i].sizeimage;
 	frm->fmt = rga_fmt_find(pix_fmt->pixelformat);
-	frm->stride = pix_fmt->bytesperline;
+	frm->stride = pix_fmt->plane_fmt[0].bytesperline;
 	frm->colorspace = pix_fmt->colorspace;
 
 	/* Reset crop settings */
@@ -539,6 +530,13 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 		  &frm->fmt->fourcc, frm->width, frm->height,
 		  frm->stride, frm->size);
 
+	for (i = 0; i < pix_fmt->num_planes; i++) {
+		v4l2_dbg(debug, 1, &rga->v4l2_dev,
+			 "plane[%d]: size %d, bytesperline %d\n",
+			 i, pix_fmt->plane_fmt[i].sizeimage,
+			 pix_fmt->plane_fmt[i].bytesperline);
+	}
+
 	return 0;
 }
 
@@ -649,14 +647,14 @@ static const struct v4l2_ioctl_ops rga_ioctl_ops = {
 	.vidioc_querycap = vidioc_querycap,
 
 	.vidioc_enum_fmt_vid_cap = vidioc_enum_fmt,
-	.vidioc_g_fmt_vid_cap = vidioc_g_fmt,
-	.vidioc_try_fmt_vid_cap = vidioc_try_fmt,
-	.vidioc_s_fmt_vid_cap = vidioc_s_fmt,
+	.vidioc_g_fmt_vid_cap_mplane = vidioc_g_fmt,
+	.vidioc_try_fmt_vid_cap_mplane = vidioc_try_fmt,
+	.vidioc_s_fmt_vid_cap_mplane = vidioc_s_fmt,
 
 	.vidioc_enum_fmt_vid_out = vidioc_enum_fmt,
-	.vidioc_g_fmt_vid_out = vidioc_g_fmt,
-	.vidioc_try_fmt_vid_out = vidioc_try_fmt,
-	.vidioc_s_fmt_vid_out = vidioc_s_fmt,
+	.vidioc_g_fmt_vid_out_mplane = vidioc_g_fmt,
+	.vidioc_try_fmt_vid_out_mplane = vidioc_try_fmt,
+	.vidioc_s_fmt_vid_out_mplane = vidioc_s_fmt,
 
 	.vidioc_reqbufs = v4l2_m2m_ioctl_reqbufs,
 	.vidioc_querybuf = v4l2_m2m_ioctl_querybuf,
@@ -683,7 +681,7 @@ static const struct video_device rga_videodev = {
 	.minor = -1,
 	.release = video_device_release,
 	.vfl_dir = VFL_DIR_M2M,
-	.device_caps = V4L2_CAP_VIDEO_M2M | V4L2_CAP_STREAMING,
+	.device_caps = V4L2_CAP_VIDEO_M2M_MPLANE | V4L2_CAP_STREAMING,
 };
 
 static int rga_enable_clocks(struct rockchip_rga *rga)

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 12/13] media: rockchip: rga: rework buffer handling for multi-planar formats

[Michael Tretter]

Multi-planar formats may have multiple planes that must be handled and
correctly mapped into a continuous buffer for the RGA by using the DMA
descriptors.

The plane offsets in the continuous mapping may now start at page
boundaries and the previous calculation based on the frame sizes is only
valid for planar buffers in a single memory. Therefore, the offsets must
be detected and set while creating the mapping.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2:
- Add size check for descriptor table when mapping video buffer
---
 drivers/media/platform/rockchip/rga/rga-buf.c | 61 ++++++++++++++++++++++-----
 drivers/media/platform/rockchip/rga/rga.c     |  7 +++
 drivers/media/platform/rockchip/rga/rga.h     |  1 +
 3 files changed, 58 insertions(+), 11 deletions(-)

diff --git a/drivers/media/platform/rockchip/rga/rga-buf.c b/drivers/media/platform/rockchip/rga/rga-buf.c
index d6774a844d9b..662c81b6d0b5 100644
--- a/drivers/media/platform/rockchip/rga/rga-buf.c
+++ b/drivers/media/platform/rockchip/rga/rga-buf.c
@@ -7,6 +7,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/scatterlist.h>
 
+#include <media/v4l2-common.h>
 #include <media/v4l2-device.h>
 #include <media/v4l2-ioctl.h>
 #include <media/v4l2-mem2mem.h>
@@ -16,7 +17,8 @@
 #include "rga-hw.h"
 #include "rga.h"
 
-static size_t fill_descriptors(struct rga_dma_desc *desc, struct sg_table *sgt)
+static ssize_t fill_descriptors(struct rga_dma_desc *desc, size_t max_desc,
+				struct sg_table *sgt)
 {
 	struct sg_dma_page_iter iter;
 	struct rga_dma_desc *tmp = desc;
@@ -24,6 +26,8 @@ static size_t fill_descriptors(struct rga_dma_desc *desc, struct sg_table *sgt)
 	dma_addr_t addr;
 
 	for_each_sgtable_dma_page(sgt, &iter, 0) {
+		if (n_desc > max_desc)
+			return -EINVAL;
 		addr = sg_page_iter_dma_address(&iter);
 		tmp->addr = lower_32_bits(addr);
 		tmp++;
@@ -40,15 +44,29 @@ rga_queue_setup(struct vb2_queue *vq,
 {
 	struct rga_ctx *ctx = vb2_get_drv_priv(vq);
 	struct rga_frame *f = rga_get_frame(ctx, vq->type);
+	const struct v4l2_pix_format_mplane *pix_fmt;
+	int i;
 
 	if (IS_ERR(f))
 		return PTR_ERR(f);
 
-	if (*nplanes)
-		return sizes[0] < f->size ? -EINVAL : 0;
+	pix_fmt = &f->pix;
 
-	sizes[0] = f->size;
-	*nplanes = 1;
+	if (*nplanes) {
+		if (*nplanes != pix_fmt->num_planes)
+			return -EINVAL;
+
+		for (i = 0; i < pix_fmt->num_planes; i++)
+			if (sizes[i] < pix_fmt->plane_fmt[i].sizeimage)
+				return -EINVAL;
+
+		return 0;
+	}
+
+	*nplanes = pix_fmt->num_planes;
+
+	for (i = 0; i < pix_fmt->num_planes; i++)
+		sizes[i] = pix_fmt->plane_fmt[i].sizeimage;
 
 	return 0;
 }
@@ -92,18 +110,39 @@ static int rga_buf_prepare(struct vb2_buffer *vb)
 	struct rga_vb_buffer *rbuf = vb_to_rga(vbuf);
 	struct rga_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
 	struct rga_frame *f = rga_get_frame(ctx, vb->vb2_queue->type);
+	ssize_t n_desc = 0;
+	size_t curr_desc = 0;
+	int i;
+	const struct v4l2_format_info *info;
+	unsigned int offsets[VIDEO_MAX_PLANES];
 
 	if (IS_ERR(f))
 		return PTR_ERR(f);
 
-	vb2_set_plane_payload(vb, 0, f->size);
+	for (i = 0; i < vb->num_planes; i++) {
+		vb2_set_plane_payload(vb, i, f->pix.plane_fmt[i].sizeimage);
+
+		/* Create local MMU table for RGA */
+		n_desc = fill_descriptors(&rbuf->dma_desc[curr_desc],
+					  rbuf->n_desc - curr_desc,
+					  vb2_dma_sg_plane_desc(vb, i));
+		if (n_desc < 0) {
+			v4l2_err(&ctx->rga->v4l2_dev,
+				 "Failed to map video buffer to RGA\n");
+			return n_desc;
+		}
+		offsets[i] = curr_desc << PAGE_SHIFT;
+		curr_desc += n_desc;
+	}
 
-	/* Create local MMU table for RGA */
-	fill_descriptors(rbuf->dma_desc, vb2_dma_sg_plane_desc(vb, 0));
+	/* Fill the remaining planes */
+	info = v4l2_format_info(f->fmt->fourcc);
+	for (i = info->mem_planes; i < info->comp_planes; i++)
+		offsets[i] = get_plane_offset(f, i);
 
-	rbuf->offset.y_off = get_plane_offset(f, 0);
-	rbuf->offset.u_off = get_plane_offset(f, 1);
-	rbuf->offset.v_off = get_plane_offset(f, 2);
+	rbuf->offset.y_off = offsets[0];
+	rbuf->offset.u_off = offsets[1];
+	rbuf->offset.v_off = offsets[2];
 
 	return 0;
 }
diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index f549966c3649..57d845c57d7d 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -365,6 +365,11 @@ static int rga_open(struct file *file)
 	ctx->in = def_frame;
 	ctx->out = def_frame;
 
+	v4l2_fill_pixfmt_mp(&ctx->in.pix,
+			    ctx->in.fmt->fourcc, ctx->out.width, ctx->out.height);
+	v4l2_fill_pixfmt_mp(&ctx->out.pix,
+			    ctx->out.fmt->fourcc, ctx->out.width, ctx->out.height);
+
 	if (mutex_lock_interruptible(&rga->mutex)) {
 		kfree(ctx);
 		return -ERESTARTSYS;
@@ -524,6 +529,8 @@ static int vidioc_s_fmt(struct file *file, void *prv, struct v4l2_format *f)
 	frm->crop.width = frm->width;
 	frm->crop.height = frm->height;
 
+	frm->pix = *pix_fmt;
+
 	v4l2_dbg(debug, 1, &rga->v4l2_dev,
 		 "[%s] fmt - %p4cc %dx%d (stride %d, sizeimage %d)\n",
 		  V4L2_TYPE_IS_OUTPUT(f->type) ? "OUTPUT" : "CAPTURE",
diff --git a/drivers/media/platform/rockchip/rga/rga.h b/drivers/media/platform/rockchip/rga/rga.h
index f3953a07d9e7..3502dff6055c 100644
--- a/drivers/media/platform/rockchip/rga/rga.h
+++ b/drivers/media/platform/rockchip/rga/rga.h
@@ -34,6 +34,7 @@ struct rga_frame {
 
 	/* Image format */
 	struct rga_fmt *fmt;
+	struct v4l2_pix_format_mplane pix;
 
 	/* Variables that can calculated once and reused */
 	u32 stride;

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 13/13] media: rockchip: rga: add NV12M support

[Michael Tretter]

Add support for the multi-planar variants of NV12. The RGA is now able
to exchange DMABUFs with other devices that only support multi-planar
NV12, for example the Hantro JPEG encoder.

Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
---
Changes in v2: None
---
 drivers/media/platform/rockchip/rga/rga.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/media/platform/rockchip/rga/rga.c b/drivers/media/platform/rockchip/rga/rga.c
index 57d845c57d7d..00fdfa9e10bc 100644
--- a/drivers/media/platform/rockchip/rga/rga.c
+++ b/drivers/media/platform/rockchip/rga/rga.c
@@ -280,6 +280,15 @@ static struct rga_fmt formats[] = {
 		.y_div = 2,
 		.x_div = 1,
 	},
+	{
+		.fourcc = V4L2_PIX_FMT_NV12M,
+		.color_swap = RGA_COLOR_NONE_SWAP,
+		.hw_format = RGA_COLOR_FMT_YUV420SP,
+		.depth = 12,
+		.uv_factor = 4,
+		.y_div = 2,
+		.x_div = 1,
+	},
 	{
 		.fourcc = V4L2_PIX_FMT_NV16,
 		.color_swap = RGA_COLOR_NONE_SWAP,

-- 
2.39.2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 00/13] media: rockchip: rga: add support for multi-planar formats

[Michael Tretter]

On Fri, 13 Oct 2023 13:00:21 +0200, Michael Tretter wrote:
> This is v2 of the series that adds support for the V4L2 multi-planar API
> to the Rockchip RGA driver. Once the RGA driver supports the
> multi-planar API, it is easier to share buffers with other V4L2 drivers
> that also support the multi-planar API and may not expose planar formats
> with contiguous planes.
> 
> v2 fixes the smatch warnings and compile errors of v1. Furthermore, now
> the DMA mask is set to 32 bits for coherent, too, and the gfp_flags are
> configured to ensure that buffers are allocated in the lower 4GB area.

Gentle Ping.

Michael

> 
> With non-contiguous planes, the U and V planes may not start at the same
> offset as with the continuous planes. Therefore, the RGA driver cannot
> rely on its calculation of the plane offsets based on the format and
> frame size anymore, but must remember the offsets when it created the
> mapping. Therefore, I also reworked how the DMA mapping is handled.
> 
> As a bonus, the RGA driver should now work correctly on devices with
> more than 4 GB of memory. Video buffers should now be allocated within
> the 4 GB boundary and an import of buffers that have higher addresses
> into the driver should fail.
> 
> Patch 1 fixes the swizzling of RGA formats. While testing all formats of
> the driver on rk3568, I discovered that the color channels of the RGB
> formats are wrong when converting to NV12. I didn't test this on other
> SoCs with an RGA and I am not sure, if they behave differently regarding
> the color channels. Please report, if this breaks the color conversion
> on other SoCs, and I will make this SoC-specific.
> 
> Patches 2 to 6 are the rework the DMA descriptor handling for the RGA
> MMU. The patches clean up, how the driver uses the DMA API, and make the
> creation of the descriptor list more explicit. Furthermore, the driver
> is changed to keep the mapping per video buffer instead of using a
> single mapping that is updated with every buffer.
> 
> Patches 7 to 11 prepare the driver for the multi-planar API including a
> cleanup of the format handling in the buffer, and finally switch to the
> multi-planar API.
> 
> Patch 12 updates the code that creates the DMA-descriptor mapping to
> correctly handle buffers with multiple planes. The driver has to iterate
> all planes and make them the continuous for the RGA.
> 
> Patch 13 enables the NV12M format, which is the multi-planar variant of
> the NV12 format.
> 
> Michael
> 
> Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
> ---
> Changes in v2:
> - Fix smatch warnings
> - Fix cast to dst_mmu_pages/src_mmu_pages to fix compile error in Patch 2
> - Remove check for upper_32_bits when filling the DMA descriptor table
> - Remove useless dma_sync_single_for_device()
> - Set DMA mask for DMA coherent
> - Set gfp_flags to __GFP_DMA32
> - Link to v1: https://lore.kernel.org/r/20230914-rockchip-rga-multiplanar-v1-0-abfd77260ae3@pengutronix.de
> 
> ---
> Michael Tretter (13):
>       media: rockchip: rga: fix swizzling for RGB formats
>       media: rockchip: rga: extract helper to fill descriptors
>       media: rockchip: rga: allocate DMA descriptors per buffer
>       media: rockchip: rga: split src and dst buffer setup
>       media: rockchip: rga: pre-calculate plane offsets
>       media: rockchip: rga: set dma mask to 32 bits
>       media: rockchip: rga: use clamp() to clamp size to limits
>       media: rockchip: rga: use pixelformat to find format
>       media: rockchip: rga: add local variable for pix_format
>       media: rockchip: rga: use macros for testing buffer type
>       media: rockchip: rga: switch to multi-planar API
>       media: rockchip: rga: rework buffer handling for multi-planar formats
>       media: rockchip: rga: add NV12M support
> 
>  drivers/media/platform/rockchip/rga/rga-buf.c | 162 ++++++++++++++++------
>  drivers/media/platform/rockchip/rga/rga-hw.c  | 146 ++++++++++++--------
>  drivers/media/platform/rockchip/rga/rga.c     | 189 ++++++++++++--------------
>  drivers/media/platform/rockchip/rga/rga.h     |  35 ++++-
>  4 files changed, 328 insertions(+), 204 deletions(-)
> ---
> base-commit: 2c1bae27df787c9535e48cc27bbd11c3c3e0a235
> change-id: 20230914-rockchip-rga-multiplanar-0d7c79b1ba93
> 
> Best regards,
> -- 
> Michael Tretter <m.tretter@pengutronix.de>
> 
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 00/13] media: rockchip: rga: add support for multi-planar formats

[Hans Verkuil]

Hi Michael,

On 13/10/2023 13:00, Michael Tretter wrote:
> This is v2 of the series that adds support for the V4L2 multi-planar API
> to the Rockchip RGA driver. Once the RGA driver supports the
> multi-planar API, it is easier to share buffers with other V4L2 drivers
> that also support the multi-planar API and may not expose planar formats
> with contiguous planes.
> 
> v2 fixes the smatch warnings and compile errors of v1. Furthermore, now
> the DMA mask is set to 32 bits for coherent, too, and the gfp_flags are
> configured to ensure that buffers are allocated in the lower 4GB area.
> 
> With non-contiguous planes, the U and V planes may not start at the same
> offset as with the continuous planes. Therefore, the RGA driver cannot
> rely on its calculation of the plane offsets based on the format and
> frame size anymore, but must remember the offsets when it created the
> mapping. Therefore, I also reworked how the DMA mapping is handled.
> 
> As a bonus, the RGA driver should now work correctly on devices with
> more than 4 GB of memory. Video buffers should now be allocated within
> the 4 GB boundary and an import of buffers that have higher addresses
> into the driver should fail.
> 
> Patch 1 fixes the swizzling of RGA formats. While testing all formats of
> the driver on rk3568, I discovered that the color channels of the RGB
> formats are wrong when converting to NV12. I didn't test this on other
> SoCs with an RGA and I am not sure, if they behave differently regarding
> the color channels. Please report, if this breaks the color conversion
> on other SoCs, and I will make this SoC-specific.
> 
> Patches 2 to 6 are the rework the DMA descriptor handling for the RGA
> MMU. The patches clean up, how the driver uses the DMA API, and make the
> creation of the descriptor list more explicit. Furthermore, the driver
> is changed to keep the mapping per video buffer instead of using a
> single mapping that is updated with every buffer.
> 
> Patches 7 to 11 prepare the driver for the multi-planar API including a
> cleanup of the format handling in the buffer, and finally switch to the
> multi-planar API.

Switching over to the multi-planar API is always a bit scary. How did you
test this? Using gstreamer? Other apps?

Regards,

	Hans

> 
> Patch 12 updates the code that creates the DMA-descriptor mapping to
> correctly handle buffers with multiple planes. The driver has to iterate
> all planes and make them the continuous for the RGA.
> 
> Patch 13 enables the NV12M format, which is the multi-planar variant of
> the NV12 format.
> 
> Michael
> 
> Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
> ---
> Changes in v2:
> - Fix smatch warnings
> - Fix cast to dst_mmu_pages/src_mmu_pages to fix compile error in Patch 2
> - Remove check for upper_32_bits when filling the DMA descriptor table
> - Remove useless dma_sync_single_for_device()
> - Set DMA mask for DMA coherent
> - Set gfp_flags to __GFP_DMA32
> - Link to v1: https://lore.kernel.org/r/20230914-rockchip-rga-multiplanar-v1-0-abfd77260ae3@pengutronix.de
> 
> ---
> Michael Tretter (13):
>       media: rockchip: rga: fix swizzling for RGB formats
>       media: rockchip: rga: extract helper to fill descriptors
>       media: rockchip: rga: allocate DMA descriptors per buffer
>       media: rockchip: rga: split src and dst buffer setup
>       media: rockchip: rga: pre-calculate plane offsets
>       media: rockchip: rga: set dma mask to 32 bits
>       media: rockchip: rga: use clamp() to clamp size to limits
>       media: rockchip: rga: use pixelformat to find format
>       media: rockchip: rga: add local variable for pix_format
>       media: rockchip: rga: use macros for testing buffer type
>       media: rockchip: rga: switch to multi-planar API
>       media: rockchip: rga: rework buffer handling for multi-planar formats
>       media: rockchip: rga: add NV12M support
> 
>  drivers/media/platform/rockchip/rga/rga-buf.c | 162 ++++++++++++++++------
>  drivers/media/platform/rockchip/rga/rga-hw.c  | 146 ++++++++++++--------
>  drivers/media/platform/rockchip/rga/rga.c     | 189 ++++++++++++--------------
>  drivers/media/platform/rockchip/rga/rga.h     |  35 ++++-
>  4 files changed, 328 insertions(+), 204 deletions(-)
> ---
> base-commit: 2c1bae27df787c9535e48cc27bbd11c3c3e0a235
> change-id: 20230914-rockchip-rga-multiplanar-0d7c79b1ba93
> 
> Best regards,


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 00/13] media: rockchip: rga: add support for multi-planar formats

[Michael Tretter]

Hi Hans,

On Fri, 24 Nov 2023 14:39:37 +0100, Hans Verkuil wrote:
> On 13/10/2023 13:00, Michael Tretter wrote:
> > This is v2 of the series that adds support for the V4L2 multi-planar API
> > to the Rockchip RGA driver. Once the RGA driver supports the
> > multi-planar API, it is easier to share buffers with other V4L2 drivers
> > that also support the multi-planar API and may not expose planar formats
> > with contiguous planes.
> > 
> > v2 fixes the smatch warnings and compile errors of v1. Furthermore, now
> > the DMA mask is set to 32 bits for coherent, too, and the gfp_flags are
> > configured to ensure that buffers are allocated in the lower 4GB area.
> > 
> > With non-contiguous planes, the U and V planes may not start at the same
> > offset as with the continuous planes. Therefore, the RGA driver cannot
> > rely on its calculation of the plane offsets based on the format and
> > frame size anymore, but must remember the offsets when it created the
> > mapping. Therefore, I also reworked how the DMA mapping is handled.
> > 
> > As a bonus, the RGA driver should now work correctly on devices with
> > more than 4 GB of memory. Video buffers should now be allocated within
> > the 4 GB boundary and an import of buffers that have higher addresses
> > into the driver should fail.
> > 
> > Patch 1 fixes the swizzling of RGA formats. While testing all formats of
> > the driver on rk3568, I discovered that the color channels of the RGB
> > formats are wrong when converting to NV12. I didn't test this on other
> > SoCs with an RGA and I am not sure, if they behave differently regarding
> > the color channels. Please report, if this breaks the color conversion
> > on other SoCs, and I will make this SoC-specific.
> > 
> > Patches 2 to 6 are the rework the DMA descriptor handling for the RGA
> > MMU. The patches clean up, how the driver uses the DMA API, and make the
> > creation of the descriptor list more explicit. Furthermore, the driver
> > is changed to keep the mapping per video buffer instead of using a
> > single mapping that is updated with every buffer.
> > 
> > Patches 7 to 11 prepare the driver for the multi-planar API including a
> > cleanup of the format handling in the buffer, and finally switch to the
> > multi-planar API.
> 
> Switching over to the multi-planar API is always a bit scary. How did you
> test this? Using gstreamer? Other apps?

I tested with GStreamer. My use case is using the RGA to convert RGB Buffers
to NV12 to feed them to a Hantro JPEG encoder. I also tried to separate the
preparatory work, the switch to the multi-planar API, and the addition of
actually multi-planar formats into separate patches.

Is there anything that I should take special care of or explicitly test, when
switching the driver to the multi-planar API?

Michael

> 
> Regards,
> 
> 	Hans
> 
> > 
> > Patch 12 updates the code that creates the DMA-descriptor mapping to
> > correctly handle buffers with multiple planes. The driver has to iterate
> > all planes and make them the continuous for the RGA.
> > 
> > Patch 13 enables the NV12M format, which is the multi-planar variant of
> > the NV12 format.
> > 
> > Michael
> > 
> > Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
> > ---
> > Changes in v2:
> > - Fix smatch warnings
> > - Fix cast to dst_mmu_pages/src_mmu_pages to fix compile error in Patch 2
> > - Remove check for upper_32_bits when filling the DMA descriptor table
> > - Remove useless dma_sync_single_for_device()
> > - Set DMA mask for DMA coherent
> > - Set gfp_flags to __GFP_DMA32
> > - Link to v1: https://lore.kernel.org/r/20230914-rockchip-rga-multiplanar-v1-0-abfd77260ae3@pengutronix.de
> > 
> > ---
> > Michael Tretter (13):
> >       media: rockchip: rga: fix swizzling for RGB formats
> >       media: rockchip: rga: extract helper to fill descriptors
> >       media: rockchip: rga: allocate DMA descriptors per buffer
> >       media: rockchip: rga: split src and dst buffer setup
> >       media: rockchip: rga: pre-calculate plane offsets
> >       media: rockchip: rga: set dma mask to 32 bits
> >       media: rockchip: rga: use clamp() to clamp size to limits
> >       media: rockchip: rga: use pixelformat to find format
> >       media: rockchip: rga: add local variable for pix_format
> >       media: rockchip: rga: use macros for testing buffer type
> >       media: rockchip: rga: switch to multi-planar API
> >       media: rockchip: rga: rework buffer handling for multi-planar formats
> >       media: rockchip: rga: add NV12M support
> > 
> >  drivers/media/platform/rockchip/rga/rga-buf.c | 162 ++++++++++++++++------
> >  drivers/media/platform/rockchip/rga/rga-hw.c  | 146 ++++++++++++--------
> >  drivers/media/platform/rockchip/rga/rga.c     | 189 ++++++++++++--------------
> >  drivers/media/platform/rockchip/rga/rga.h     |  35 ++++-
> >  4 files changed, 328 insertions(+), 204 deletions(-)
> > ---
> > base-commit: 2c1bae27df787c9535e48cc27bbd11c3c3e0a235
> > change-id: 20230914-rockchip-rga-multiplanar-0d7c79b1ba93
> > 
> > Best regards,
> 
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 00/13] media: rockchip: rga: add support for multi-planar formats

[Hans Verkuil]

On 24/11/2023 16:31, Michael Tretter wrote:
> Hi Hans,
> 
> On Fri, 24 Nov 2023 14:39:37 +0100, Hans Verkuil wrote:
>> On 13/10/2023 13:00, Michael Tretter wrote:
>>> This is v2 of the series that adds support for the V4L2 multi-planar API
>>> to the Rockchip RGA driver. Once the RGA driver supports the
>>> multi-planar API, it is easier to share buffers with other V4L2 drivers
>>> that also support the multi-planar API and may not expose planar formats
>>> with contiguous planes.
>>>
>>> v2 fixes the smatch warnings and compile errors of v1. Furthermore, now
>>> the DMA mask is set to 32 bits for coherent, too, and the gfp_flags are
>>> configured to ensure that buffers are allocated in the lower 4GB area.
>>>
>>> With non-contiguous planes, the U and V planes may not start at the same
>>> offset as with the continuous planes. Therefore, the RGA driver cannot
>>> rely on its calculation of the plane offsets based on the format and
>>> frame size anymore, but must remember the offsets when it created the
>>> mapping. Therefore, I also reworked how the DMA mapping is handled.
>>>
>>> As a bonus, the RGA driver should now work correctly on devices with
>>> more than 4 GB of memory. Video buffers should now be allocated within
>>> the 4 GB boundary and an import of buffers that have higher addresses
>>> into the driver should fail.
>>>
>>> Patch 1 fixes the swizzling of RGA formats. While testing all formats of
>>> the driver on rk3568, I discovered that the color channels of the RGB
>>> formats are wrong when converting to NV12. I didn't test this on other
>>> SoCs with an RGA and I am not sure, if they behave differently regarding
>>> the color channels. Please report, if this breaks the color conversion
>>> on other SoCs, and I will make this SoC-specific.
>>>
>>> Patches 2 to 6 are the rework the DMA descriptor handling for the RGA
>>> MMU. The patches clean up, how the driver uses the DMA API, and make the
>>> creation of the descriptor list more explicit. Furthermore, the driver
>>> is changed to keep the mapping per video buffer instead of using a
>>> single mapping that is updated with every buffer.
>>>
>>> Patches 7 to 11 prepare the driver for the multi-planar API including a
>>> cleanup of the format handling in the buffer, and finally switch to the
>>> multi-planar API.
>>
>> Switching over to the multi-planar API is always a bit scary. How did you
>> test this? Using gstreamer? Other apps?
> 
> I tested with GStreamer. My use case is using the RGA to convert RGB Buffers
> to NV12 to feed them to a Hantro JPEG encoder. I also tried to separate the
> preparatory work, the switch to the multi-planar API, and the addition of
> actually multi-planar formats into separate patches.
> 
> Is there anything that I should take special care of or explicitly test, when
> switching the driver to the multi-planar API?

No, I just wanted to have a better overview of how this driver is used and
how you tested it.

Regards,

	Hans

> 
> Michael
> 
>>
>> Regards,
>>
>> 	Hans
>>
>>>
>>> Patch 12 updates the code that creates the DMA-descriptor mapping to
>>> correctly handle buffers with multiple planes. The driver has to iterate
>>> all planes and make them the continuous for the RGA.
>>>
>>> Patch 13 enables the NV12M format, which is the multi-planar variant of
>>> the NV12 format.
>>>
>>> Michael
>>>
>>> Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
>>> ---
>>> Changes in v2:
>>> - Fix smatch warnings
>>> - Fix cast to dst_mmu_pages/src_mmu_pages to fix compile error in Patch 2
>>> - Remove check for upper_32_bits when filling the DMA descriptor table
>>> - Remove useless dma_sync_single_for_device()
>>> - Set DMA mask for DMA coherent
>>> - Set gfp_flags to __GFP_DMA32
>>> - Link to v1: https://lore.kernel.org/r/20230914-rockchip-rga-multiplanar-v1-0-abfd77260ae3@pengutronix.de
>>>
>>> ---
>>> Michael Tretter (13):
>>>       media: rockchip: rga: fix swizzling for RGB formats
>>>       media: rockchip: rga: extract helper to fill descriptors
>>>       media: rockchip: rga: allocate DMA descriptors per buffer
>>>       media: rockchip: rga: split src and dst buffer setup
>>>       media: rockchip: rga: pre-calculate plane offsets
>>>       media: rockchip: rga: set dma mask to 32 bits
>>>       media: rockchip: rga: use clamp() to clamp size to limits
>>>       media: rockchip: rga: use pixelformat to find format
>>>       media: rockchip: rga: add local variable for pix_format
>>>       media: rockchip: rga: use macros for testing buffer type
>>>       media: rockchip: rga: switch to multi-planar API
>>>       media: rockchip: rga: rework buffer handling for multi-planar formats
>>>       media: rockchip: rga: add NV12M support
>>>
>>>  drivers/media/platform/rockchip/rga/rga-buf.c | 162 ++++++++++++++++------
>>>  drivers/media/platform/rockchip/rga/rga-hw.c  | 146 ++++++++++++--------
>>>  drivers/media/platform/rockchip/rga/rga.c     | 189 ++++++++++++--------------
>>>  drivers/media/platform/rockchip/rga/rga.h     |  35 ++++-
>>>  4 files changed, 328 insertions(+), 204 deletions(-)
>>> ---
>>> base-commit: 2c1bae27df787c9535e48cc27bbd11c3c3e0a235
>>> change-id: 20230914-rockchip-rga-multiplanar-0d7c79b1ba93
>>>
>>> Best regards,
>>
>>


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel