On Thursday 29 August 2013 06:58 PM, Hans Verkuil wrote:
On Thu 29 August 2013 14:32:49 Archit Taneja wrote:
VPE is a block which consists of a single memory to memory path which can
perform chrominance up/down sampling, de-interlacing, scaling, and color space
conversion of raster or tiled YUV420 coplanar, YUV422 coplanar or YUV422
interleaved video formats.
We create a mem2mem driver based primarily on the mem2mem-testdev example.
The de-interlacer, scaler and color space converter are all bypassed for now
to keep the driver simple. Chroma up/down sampler blocks are implemented, so
conversion beteen different YUV formats is possible.
Each mem2mem context allocates a buffer for VPE MMR values which it will use
when it gets access to the VPE HW via the mem2mem queue, it also allocates
a VPDMA descriptor list to which configuration and data descriptors are added.
Based on the information received via v4l2 ioctls for the source and
destination queues, the driver configures the values for the MMRs, and stores
them in the buffer. There are also some VPDMA parameters like frame start and
line mode which needs to be configured, these are configured by direct register
writes via the VPDMA helper functions.
The driver's device_run() mem2mem op will add each descriptor based on how the
source and destination queues are set up for the given ctx, once the list is
prepared, it's submitted to VPDMA, these descriptors when parsed by VPDMA will
upload MMR registers, start DMA of video buffers on the various input and output
clients/ports.
When the list is parsed completely(and the DMAs on all the output ports done),
an interrupt is generated which we use to notify that the source and destination
buffers are done.
The rest of the driver is quite similar to other mem2mem drivers, we use the
multiplane v4l2 ioctls as the HW support coplanar formats.
Signed-off-by: Archit Taneja <arc...@ti.com>
Thanks for the patch. Just a few small comments below...
---
drivers/media/platform/Kconfig | 16 +
drivers/media/platform/Makefile | 2 +
drivers/media/platform/ti-vpe/Makefile | 5 +
drivers/media/platform/ti-vpe/vpe.c | 1740 ++++++++++++++++++++++++++++++
drivers/media/platform/ti-vpe/vpe_regs.h | 496 +++++++++
5 files changed, 2259 insertions(+)
create mode 100644 drivers/media/platform/ti-vpe/Makefile
create mode 100644 drivers/media/platform/ti-vpe/vpe.c
create mode 100644 drivers/media/platform/ti-vpe/vpe_regs.h
diff --git a/drivers/media/platform/ti-vpe/vpe.c
b/drivers/media/platform/ti-vpe/vpe.c
new file mode 100644
index 0000000..85b0880
--- /dev/null
+++ b/drivers/media/platform/ti-vpe/vpe.c
<snip>
+static int vpe_enum_fmt(struct file *file, void *priv,
+ struct v4l2_fmtdesc *f)
+{
+ if (V4L2_TYPE_IS_OUTPUT(f->type))
+ return __enum_fmt(f, VPE_FMT_TYPE_OUTPUT);
+ else
The line above isn't necessary.
Oh right, thanks for spotting that.
+ return __enum_fmt(f, VPE_FMT_TYPE_CAPTURE);
+}
+
<snip>
+
+ pix->field = V4L2_FIELD_NONE;
+
+ v4l_bound_align_image(&pix->width, MIN_W, MAX_W, W_ALIGN,
+ &pix->height, MIN_H, MAX_H, H_ALIGN,
+ S_ALIGN);
+
+ pix->num_planes = fmt->coplanar ? 2 : 1;
+ pix->pixelformat = fmt->fourcc;
+ pix->colorspace = fmt->fourcc == V4L2_PIX_FMT_RGB24 ?
+ V4L2_COLORSPACE_SRGB : V4L2_COLORSPACE_SMPTE170M;
pix->priv should be set to NULL as well.
I'll fix this.
<snip>
+}
+
+#define V4L2_CID_TRANS_NUM_BUFS (V4L2_CID_USER_BASE + 0x1000)
Reserve a control range for this driver in include/uapi/linux/v4l2-controls.h.
Similar to the ones already defined there.
That will ensure that controls for this driver have unique IDs.
Thanks, I took this from the mem2mem-testdev driver, a test driver
doesn't need to worry about this I suppose.
I had a query regarding this. I am planning to add a capture driver in
the future for a similar IP which can share some of the control IDs with
VPE. Is it possible for 2 different drivers to share the IDs?
Also, I noticed in the header that most drivers reserve space for 16
IDs. The current driver just has one, but there will be more custom ones
in the future. Is it fine if I reserve 16 for this driver too?
<snip>
+
+static int queue_init(void *priv, struct vb2_queue *src_vq,
+ struct vb2_queue *dst_vq)
+{
+ struct vpe_ctx *ctx = priv;
+ int ret;
+
+ memset(src_vq, 0, sizeof(*src_vq));
+ src_vq->type = V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE;
+ src_vq->io_modes = VB2_MMAP;
+ src_vq->drv_priv = ctx;
+ src_vq->buf_struct_size = sizeof(struct v4l2_m2m_buffer);
+ src_vq->ops = &vpe_qops;
+ src_vq->mem_ops = &vb2_dma_contig_memops;
+ src_vq->timestamp_type = V4L2_BUF_FLAG_TIMESTAMP_MONOTONIC;
Shouldn't this be TIMESTAMP_COPY?
Right, it should be, I'll fix it.
+
+ ret = vb2_queue_init(src_vq);
+ if (ret)
+ return ret;
+
+ memset(dst_vq, 0, sizeof(*dst_vq));
+ dst_vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE;
+ dst_vq->io_modes = VB2_MMAP;
+ dst_vq->drv_priv = ctx;
+ dst_vq->buf_struct_size = sizeof(struct v4l2_m2m_buffer);
+ dst_vq->ops = &vpe_qops;
+ dst_vq->mem_ops = &vb2_dma_contig_memops;
+ dst_vq->timestamp_type = V4L2_BUF_FLAG_TIMESTAMP_MONOTONIC;
Ditto.
Thanks for the review.
Archit
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
